Herpes Simplex Virus and Human Cytomegalovirus Replication in WI-38 Cells I. Sequence of Viral Replication

JOURNAI OFVIROLOGY., Oct. 1973,p.919-930

CopyrightO 1973 American Society forMicrobiology

Vol. 12. No. 4

Printedin U.S.A.

Herpes Simplex

Virus

and Human

Cytomegalovirus

Replication

in

WI-38

Cells

I.

Sequence

of

Viral

Replication

JANET DUYCKINCK SMITH' AND ETIENNE DE HARVEN

Sloan-Kettering Institute for Cancer Research and Cornell University Graduate School of Medical Sciences,

New

York,New York 10021

Received forpublication 26 April 1973

A comparison, under standardized conditions, of herpes simplex virus

(HSV) and human cytomegalovirus (CMV) revealeddifferences in viral

morphol-ogy, inthetimingof their infectiouscycles, and in several morphological events

during those cycles. Structuraldistinctionsbetween thetwoviruses included the

coatingofunenveloped cytoplasmic CMV capsids, but not those ofHSV, and a

variation in thestructureoftheircores. Since thetwo cycleswerecarried outin

thesame host cell strain under conditions ofone-stepgrowth (input multiplicity

= 10 PFU/cell), itwaspossibleto constructtime scaleslocating the majorevents

of each cycle. Comparison of thetwoshowed that HSVreplicated and released

progeny within 8 hpostinfection, whereas CMV required 4days. These results

correlated well with those ofconcurrent plaque assays. Within the longer CMV

cycle, most of the major events appeared retarded toa similar degree, and no

obvious limiting step in particle production could be identified. Distinctions

between thetwo cyclesincluded thefollowing: condensation of the chromatin in

HSV- but not CMV-infected cells; the greater tendency of HSV to produce

membranealterations; and theappearanceofcytoplasmicdense bodies in

CMV-but not HSV-infected cells. Identification of these differences even under

identical conditions of culture and infection strongly implies that they result from intrinsic differences in the nature of the viruses, and are not caused by variations in experimental conditions.

The members of theherpesvirus groupshare

common features including a similar

morphol-ogy, a nuclear site of capsid assembly, and similar modes of maturation and release, yet differ significantly in otherparameterssuch as

the length of the replication cycle, the host range, and theyield ofinfectious cell-freevirus (12, 17, 32). To date many of the studies of individual herpesviruses have been carried out

under different conditions in different host cell systems, making it difficult to determine

whether observed dissimilaritieswereduetothe intrinsic nature ofthe viruses or to the

varia-tions in the systems employed. We report here

on acomparison ofthereplicative cycles of two

herpesviruses carried out under standardized conditions in the same host cell system, using equivalent multiplicities ofinfection. To estab-lish a time scale for each cycle, a multiplicity

lPresent address: Department of Anatomy, The Medical

CollegeofPennsylvania. Philadelphia. Pa. 19144.

919

sufficient to produce a synchronous one-step growth cycle wasused.

MATERIALS AND METHODS

Cells. WI-38cellswereobtainedfromthe American

Type Culture Collection in the 16th passage. They

were grown as stationary monolayers in minimal

essentialmedium (MEM) with Earle's saltsplus10%

fetal calfserum (FCS), 7.7 mM NaHCO3, 100jgof

streptomycinperml, and 100Uofpotassium

penicil-lin G per ml. Tests for mycoplasma contamination

were negative.

Virus stocks. Herpes simplex virus(HSV), strain

HF (8)wasreceivedfromthe AmericanType Culture

Collection, as was human cytomegalovirus (CMV),

strain AD-169 (28). Both were free of mycoplasma.

Virus-infected cells were maintained on MEM with 2%agammacalfserum(MEM2), and were harvested

when85 to 90%showedcytopathic effects.

Mechani-calhomogenizationwasfollowedbycentrifugation at

500 x g for 10 minand filtration through prewashed

Milliporefilters with pore size of 0.8 and then 0.45jm

(6).Clarifiedsuspensions ofHSVwerefrozen

on November 10, 2019 by guest

http://jvi.asm.org/

SMITH AND DE HARVEN

ately, whereas those ofCMV were concentrated

10-to20-fold by negativepressureultrafiltration through

dialysis tubing(25). Thefrozen stocksweretitered by

plaque assay.

Plaque assay. Cell-free virus wasassayed bythe

procedures of Wentworth and French for both HSV

(29) and CMV(30), except that CMVwas dilutedin

medium lacking bicarbonate, and fixed monolayers

werestained with 1%aqueousmethyleneblue.

Infection of cell monolayers. Preconfluent

mono-layerswereinfected with0.1 ml/plateofcellfree virus

ata multiplicityof 10PFU/cell.After adsorptionfor

90minat37C and5%CO2, thecultureswerewashed

onceand incubated in 2 ml offresh MEM 2.

HSV-infected samplesweretakenat0, 2, 3,4, 5,6, 8, 10, 12,

and24 hpostinfection, with uninfected controlsat0

and 24 h. Asample of medium from each plate was

frozen and subsequently titered for released virus.

CMV-infectedcellswerefixedat12-h intervalsover a

periodof 8days,withuninfected controlsat0 and 7.5

days. Duplicate plates were takenat eachsampling

time, onefor titration of released and cell-associated

virusand the other for electron microscopy.

Electron microscopy. The monolayers werefixed

and embeddedin situbyamodification of the method

of Brinkley et al. (3). Sections were stained with

uranylacetate and lead citrate, coated withcarbon,

and examined in a modified Siemens 1A electron

microscope. The main modification was a shorter

objectivelens focal distance(H. Armbruster,Siemens

Inc., Iselin, N.J.).

Fornegative staining,unfixed viralspecimenswere

adsorbedto grids covered bythin carbon films, and

then stained with 2% aqueous phosphotungstic acid

(PTA), pH4.8.

RESULTS

Virusgrowthcurves. The growth curvesfor

thecycles studied by electron microscopy (Fig. 1) indicate that, as expected with an input

multiplicity of 10 PFU/cell, conditions of

one-step infectionwere achieved with bothviruses.

This fact was independently confirmed by

in-fectious center assays.

Viralmorphology. The major structural dif-ferences between the two viruses were as

fol-lows: (i) In negatively stained samples, the

cores ofHSV appeared to becontinuous

struc-tureswithawall of uniformthickness (Fig. 2d),

whereas those ofCMV resembledaggregates of globular subunits (Fig. 2b). This distinction in

corestructure wasalso evidentinthinsections,

where CMV but not HSV cores had a beaded appearanceincross section(Fig. 3c, e). Despite a recent report on herpes simplex morphology

(9), we were unable tofind any evidence indi-cating a toroidal structure for the core. (ii)

Unenveloped CMV capsids located free in the

cytoplasm were coated by a layer of fibrillar

material applied directly to the exterior of the capsid (Fig. 3d), whereas free cytoplasmic

cap-sids of HSV werenever coated (Fig. 3f). Coats

were never present on those capsids ofHSVor

CMV located within intact nuclei (Fig. 3e). Other features of viral morphology, including the variety of structural types usually observed with herpesviruses (25), areillustratedinFig. 2

and 3.

Comparison of paired HSV and CMV means, using Student's t test, showed no significant differences between any dimensions of the two (a = 0.1). In thin section, enveloped particles averaged 174.0 nm in diameter, capsids

aver-aged 106.4 nm, andcoresaveraged 64.3nm.The average thickness of the coat on unenveloped cytoplasmic CMV capsids was24.4 nm.

CMV replication cycle. By electron micros-copy the first visible sign of infection was the

rounding up of individual uninucleate cells

between 12and24 hpostinfection. The reniform nuclei were displaced to the periphery of the cytoplasm, while a prominent Golgi apparatus

(Fig.4) occupied the centerofthe cell.

Less frequently cells fused to produce small

polykaryocytes usuallycontaining 3 to 4nuclei.

Thesewereoftenarranged in aringsurrounding

theGolgielements.

By day 3 capsids were assembling in the nucleus. At this stage they were all of one

morphological type, namely those containing

"empty" cores. Occasionally a capsid and core

appeared to be developing in synchrony. Free cores without a surrounding capsid were not observed. Capsids with dense cores began to appear between 3 and 3.5 days.

At 3.5 days mostofthe capsidswere closely

associated with a distinctiveskein-like nuclear

inclusion (Fig. 5), although smaller numbers

were found in the clear zone of nucleoplasm which characteristically separated these

inclu-sionsfromthe nuclearmembrane.

Envelopmentatthenuclear membranebegan between3and3.5days.

Capsids

ofall morpho-logicaltypes buddedthrough the innernuclear

membrane, either directly into the perinuclear cisterna (Fig. 6), or into sac-like membranous invaginations commonly found in the nucleo-plasm (Fig. 7). Rupture of the nuclear

mem-brane was not observed attheseearly stages. In theperiodbetween2and 3days postinfec-tion, there was a transient increase in the number of 55- to 60-nm intranuclear

granules

(Fig. 8, 9). Although previously reported

by

others (23) and interpreted as

possible

viral

nucleoids, these seem morecloselyto resemble theperichromatingranulesofnormal

cells,

both

920 J. VIROL.

on November 10, 2019 by guest

http://jvi.asm.org/

HSV AND CMV REPLICATION IN WI-38 CELLS

in size and in the presence of a clear area or

"halo" separating many ofthem from hetero-chromatin.

From theperinuclearcisterna, enveloped par-ticles usually moved into the cytoplasm within smooth vacuoles, formed by the evagination of the outer nuclear membrane. Nonlytic release of viruses within vacuoles occurred when the vacuolar membrane fused with the

plasma-lemma, thereby extruding the enveloped parti-cle.Altematively, particles could leave the

peri-nuclear cisterna via the channels of the rough

endoplasmic reticulum (Fig. 10), and could

travel through these channels for variable dis-tances. Since no continuities were ever found

between the rough endoplasmic reticulum and theplasmalemma, theviruses within the

reticu-lum must exit from the cell by a discontinuous route involving intermediate vacuoles.

The effect of CMV on cellular membranes was minimal. The nuclear membrane redu-plication which is characteristic of herpes-viruses occurred only to a limited extent (Fig. 11). Reduplication seldom involved more than

one extra membrane pair or more than a short section of the nuclearmembrane. Similarly, cell fusion was not common, and those

polykaryo-cytes observed rarely had more than 3 or 4

nuclei. In the cytoplasm there was a

prolifera-tion of

Golgi

vesiclesandstackedsaccules

(Fig.

12), but these retained their normal

morpholo-gies.

Within the nuclei there was no noticeable

condensation or margination of the chromatin either at this or later stages of the infection.

Despite the apparently intact nuclear

mem-branes, unenveloped capsids were routinely

ob-served in the cytoplasm beginning at 4 days.

These capsids, all of which were coated, were mostnumerous in theimmediate vicinity of the nucleus and in theGolgiregion (Fig. 12e). In the

Golgi

zone many budded into a variety of

cytoplasmic vacuoles including multivesicular bodies,

phagolysome-like

vacuoles, and a char-acteristic small, clear vacuole with a

crescent-shapedcrosssection(Figs. 12a-c).As aresult of this process the capsids simultaneously

ac-quired an envelope and became enclosed in a

surrounding vacuole, making them virtually identical to viruses enveloped at the nuclear membrane.

Concurrent with capsid assembly and

en-velopment, a homogenous electron-dense mate-rial began to accumulate in the cytoplasm, especially in the Golgi region (Fig. 12e). It

appeared first in small, scattered foci which lacked a limiting membrane. These retained

iolO

10f1

10 4

1o6~ L-

0-

1-10

102

k

10oi

2 6 10

HoursPost Infection

16

~

U-a- 104

102

- Released virus

n-n,Cell-associated virus

* nocula

1 3 5 7 9 11

0

Days Post Infection

FIG. 1. One-step growth curves showing the titers

of released virus for HSV (upper curve), and both

released and cell-associated virus for CMV (lower

curve).

regular spherical outlines while gradually in-creasing in size. By day 5 many foci were acquiring anenvelopebybuddingintothe same types ofvacuolesas CMV capsids (Fig. 12d, e).

Individual vacuoles often contained a mixed populationofboththeseenvelopeddense bodies and enveloped CMV. The dense bodies were

released from the cell in the same manner as virus.

In later stages, a variety of distinctive

nu-clear inclusions were observed

(Fig.

13, 14).

These may represent sequential stages in a

gen-eral compaction or condensation ofthe

skein-like inclusions described above.

Cell lysis commonly occurred at 7 to8 days postinfection.

HSV replication cycle. The most obvious dissimilarity between the cycles of HSV and

CMV was in the length of time necessary for

replication (Fig. 1). A comparison of the timing ofthemajor events in the two cycles is shown in Table 1.

Although the replication times differ signifi-cantly, most of the major events occur in the same sequence in both cycles. The exceptions 921

VOL.12,1973

on November 10, 2019 by guest

http://jvi.asm.org/

[image:3.495.262.432.74.342.2]

SMITH AND DE HARVEN

2a

1

'IO. 1~ .

2--b

t

.:

2d

2c

3a

3b

(I

--W~~~~~~~~~~.f

. .;

F.

. , r I

*i.s

;*

c

FIG. 2. Negatively stained HSV and CMV particles. (a) Short 12- to 15-nm projections (arrowheads)

sometimes seen on the outer surface of enveloped particles. HSV. (b) Enveloped CMV with individual

capsomeresvisibleattheperiphery of the capsid (arrow). The core appears as an aggregate of globular subunits

(arrowhead),sixormoreperipheral and one central. (c) HSVparticle showing the hollow polygonal shape of the

capsomeres seen end-on. (d)AnHSV capsid containing a central core made up of two continuous concentric

structures (arrows). (A magnification ofx127,000 was used for a-d.)

FIG. 3. HSV and CMV particlesinthinsection. (a) Enveloped HSV. Note the dense material marginated

against the inner surface of the envelope (arrowhead), the capsid, and an apparently ruptured core of intermediate electron density. (b) An unsual HSV capsid completely filled by the electron dense material

thought to represent viral nucleoprotein. (c)HSV capsids containing "empty" cores. Note that as in Fig. 2d, the

wallof thecore seemscomposed oftwolayers, theinner here appearing slightly more electron opaque. (d) An

unenveloped, cytoplasmic CMV capsid, with fibrillar coat. Traces ofa coreremain intheinterior,along with an

unknown materialforminganunusual tail-like structure (arrowhead). (e) ACMV capsid within the nucleus.

Note the absenceof any coat and the beaded appearance of the core. (f)A cytoplasmic HSV capsid lacking a

core entirely and which, like all HSV capsids, nuclear or cytoplasmic, is not coated. (A magnification of

approximately x 127,000 wasusedfora-f.)

922 J. VIROL.

;.s

:4.

on November 10, 2019 by guest

http://jvi.asm.org/

[image:4.495.112.406.78.493.2]

HSV AND CMV REPLICATION IN WI-38 CELLS

r

*-.

:,.4;,;,>;:>s'

'II. *1

f '0

:

*'^; - ^

wf ...

.

b ,# F '. s

>,,@ .1 :, ,\

tr/) fE.

( "-e

.s , Z.l, 's. f

es

F

Y

t.t

s t ,

'--^ - -X}t* \ ' I_.j%i,.

+lf -- s Kbi\t r

., :;

.. ..

-v

<. .. s

87

-.1 N I

.: "..I

.4'.

4

6

*''

'v{'4

*'*

'

'.;

a

(_

.:S

s le'% ' j -v~~~%

FIG. 4. A rounded cell 12h after CMVinfection. To the right of the reniform nucleus is aregion ofthe

cytoplasm where the accumulationand proliferationofGolgi elements have begun. Magnification x6,700.

FIG. 5. A typical nuclearinclusion3.5daysafter CMV infection.Itis composedofirregularanastomosing

strandsof electron dense materialand is in close proximity to mostof thedeveloping capsids.Magnification

x13,000.

FIG. 6. Asingle CMVcapsid acquiringanenvelope by buddingthroughthe inner nuclear membrane intothe

perinuclearcisterna. Three and one-halfdays postinfection. Magnification x37,000.

FIG. 7. CMVcapsidenvelopmentby buddingintoamembrane-boundedsacwhich ispresumablycontinuous

withtheperinuclearcisterna. Fourdays postinfection. Magnificationx25,500.

VOL.12,1973 923

-;.r,

.11

fz

on November 10, 2019 by guest

http://jvi.asm.org/

[image:5.495.54.435.80.550.2]

SMITH AND DE HARVEN

pi.

., . . ,

.49

144

. X

x

r>* ;

_ t.;

Ab.. ;

'q-,.*

S

!

~ rf

rAr

AA'tf.

4

:V

k .r*

_,p

1||

f0,,Y.

9

FIG. 8. A large accumulation of "perichromatin-like" granules inanHSV-infected nucleus. The granules occur in association with heterochromatin and occasionally near characteristic granular aggregates. 2 h postinfection. Magnification x36,000.

FIG. 9. "Perichromatin-like"granules inanucleus 2days after infection with CMV. Note the clear regionor

"halo"separating them from the surrounding heterochromatin. Magnification x38,000.

FIG. 10. Atangential section ofanenveloped particle in theperinuclear cisternanearthepoint of continuity

with therough endoplasmic reticulum. HSV, 8hpostinfection. Magnification x75,000.

FIG. 11. At severalpoints (arrowheads) short segments of reduplicated membranes separate the nucleus

(lower right) from the cytoplasm ofacell 4days after infectionwith CMV.Magnification x48,000.

924 J.VIROL.

I

I _

on November 10, 2019 by guest

http://jvi.asm.org/

[image:6.495.67.447.74.552.2]

.12.a.

t.

ze rr--n x

12b

12c.-,¢;

r,,^

A..

... ,

..I

¢2r -?

fr..'.:."

*; 1

4.

,t ir t*C., g 2,

*Xso;

sr,

f ib

.sOX

Xs s

S..

ffi X

#,,.. r e

,.;

Xt_

.. i'

.I ~~~~.

-. i ..

.-, x

A-z

'-

1>sr,

A

*. ;iXZ [.4: A

PYtS'

FIG. 12. Cytoplasmic envelopmentinCMVinfection. (a-c)NakedCMVcapsids budding throughonesideof

acytoplasmicvacuole. Note that inbudding particlesthe thicknessofthecapsidcoatissignificantlyreduced.

Magnification approximately x100,000. (d)A typical dense bodyundergoing cytoplasmic envelopment in a mannersimilartoviralcapsids. Six dayspostinfection. Magnification x32,000. (e)Anareaof cytoplasmrichin

Golgi elements and containing several unenveloped, coated capsids (arrows, upper left), one of which

(arrowhead)isbuddingintoasmallcytoplasmicvacuole.Manydensebodies,bothenvelopedandunenveloped

(arrow,lowerright)arealso present. Four andone-halfdays postinfection. Magnification x24,000.

925

~~rv

12 d

--*4

.4

.*W

so ,@.

, v ;t

> e

.- i: vi_

s 7 .t

,

*!^

.

._-

c :

o

O

' .. .:1:

i 'W

4

..

.4 It,

i.

12

e

s.,

.".1.4

-V.

%. i!

ll.,.-. I

i .

-. , .1.

I'l.l-*

on November 10, 2019 by guest

http://jvi.asm.org/

SMITH AND DE HARVEN

TABLE 1. Comparison of timing of major events in

HSV and CMV cycles.

Onset Event

CMV(days) HSV(h)

Cell fusion androunding 0.5 6

Golgialteration 1 8

Viral eclipse 2 2

Appearance of increased 2 3

numbersof

perichromatin-like granules

Condensationofthe chroma- Not ob- 3

tin served

Assembly of the first capsids 3 4

inthenucleus

Appearance of the first cap- 3.5 5

sids with dense cores

Envelopment at the nuclear 3.5 6

membrane

Appearance of naked capsids 4 6

inthecytoplasm

Appearance of dense cyto- 4 Not

ob-plasmic aggregates served

Firstreleased particles 4 8

Nuclear membrane redu- 4 8

plication

Envelopment at cytoplasmic 4 8

membranes

Celllysis 7-8 24-48

arethose reflecting thereorganization of

cellu-lar architecture. These, including cell fusion, cellrounding, and the proliferation and consoli-dation of theGolgi elements,occurwithinhours

after infection by either virus, and therefore

relativelyearlier in the longer cycle of CMV. Inadditiontothesetemporal variations there

werequalitative differences whichmaybe

sum-marized asfollows:

(i) The naked cytoplasmic capsids of HSV,

unlike thoseofCMV, were nevercoated.

(ii) The cytoplasmic dense bodies (Fig. 12e) characteristic of CMV infectionwerenotfound

inHSV-infected cells.

(iii) The skein-like nuclear inclusions of

CMV-infected cells were not observed after

infection with HSV. Nuclear inclusions were

not commonuntil late inthereplicationcycle,

when two characteristic types appeared (Fig. 15, 16).

(iv)By3h postinfection with HSV therewas a marked condensation and marginationof the

chromatin, which persistedthroughout replica-tion (Fig. 17, 18). Such condensation was

en-tirely absentwith CMV.

(iv) By3 h postinfectionwithHSV therewas

a marked condensation and margination of the

chromatin, which persistedthrough replication (Fig. 17, 18). Such condensation was entirely

absent with CMV.

(v) Extensive membrane changes followed

infection bythis strain ofHSV. These included

the following: (i) cell fusion with formation of

large polykaryocytes often containing 12 or

more nuclei; (ii) reduplication of the nuclear membrane, with two or more extra membrane pairs over much of the periphery of thehighly lobulated nucleus; and (iii) proliferation and

extensive morphological alterationof the

mem-branous elements in the Golgi region. With respect to this last point, by 8 h postinfection

the typical vesicles and saccules of the Golgi

had been almost entirely replaced by two new

elements, namely, concentric cisternae

bounded by smooth membranes (Fig. 18) and

large smooth-surfaced vacuoles arranged in

characteristic clusters (Fig. 17, 18).

Cytoplas-mic envelopment ofnaked HSV capsids often

occurred attheselarge vacuoles.

DISCUSSION

Since both viral cycles described above were carried out under standardized conditions, the observed differences between the two presum-ably reflectdifferencesinthe geneticmakeupof the viruses rather than variation inthe experi-mental system.

To the extent that other strains of HSV or CMV are related to those used here, they should showsimilar distinctive patterns of replication when compared inthe same cell system. Indeed, earlier investigations with other strains have shown similar patternswith regard to chroma-tinmargination andcondensation byHSV(15); lack of condensation with CMV (5, 10, 13, 18); extensive membrane changes with HSV, i.e., eitherpolykaryocytosisorextensive cell round-ing (12, 16); and onlyminimaleffects with CMV (13, 18). Nor are the patterns ofreplicationwe observed limited to laboratory strains. Thus, extensive polykaryocytosis has been observed notonly with adaptedstrains ofHSV, but also with recently isolated virus (1, 7, 22), andeven inbiopsy material(2, 20,21).Asimilar concord-ance between many of the major effects of an adapted strain and afreshisolatewasreported

recentlyforCMV (11).

Withregardtoviralmorphology,weobserved

many of the same featuresreportedby othersfor

HSV (15, 31) and CMV (13, 26). The

double-walledstructure oftheHSVcoreresembles that describedfortheLucke virus (27), whereas the beaded orglobularappearance of theCMVcore has been illustrated by a number ofpublished micrographs (5, 10, 11, 13). We did not find evidence witheither HSVorCMV foratoroidal structure ofthe core, suchasreportedrecently

for HSV(9).Thiswas true eveninlatestages of

926 J. VIROL.

on November 10, 2019 by guest

http://jvi.asm.org/

[image:8.495.58.251.89.352.2]

0

' X

-;1I-e

4'-*

I

4.

13

.'

2.

.\.

.

4:

#

w

el

ti/eA~-t;t

.t--<-7

FIG. 13. A late CMV inclusioncomposed of multiple roundmasses,associatedattheirperiphery with viral

capsids andwith theremnantsof theearlierskein-like inclusions,andoccupyingan areaof reduced electron

opacity inthecenterof thenucleus. A rim of nucleoplasmseparatestheinclusionfromthe nuclear membrane andfromthesegregated nucleolusatthe lowerright. Six dayspostinfection. Magnification x6,300.

FIG. 14. Several inclusionsresembling granulesorshortfibrils (arrow) andone morecompactfibrillar form

(arrowhead) are closely associated with the remnants of the skein-like CMV nuclear inclusion. Six days

postinfection.Magnification x18,000.

FIG. 15. A late HSV inclusion ispresent in the lowerright. Note also the extensive reduplication ofthe

nuclear membrane as well as the viruses acquiring an envelope at their inner face (arrow). Twelve hours

postinfection.Magnification x34,000.

FIG. 16. Around inclusionbody (upperright)commoninlateHSVinfection.Thesewerescatteredsinglyin

thenucleoplasmrather thaningroups,andshowednotraceofafibrillarsubstructure. Note also thenumerous

small, ring-likestructuresfreeinthenucleoplasm (arrow). Althoughsmaller thannormal viralcores,several

appearto besurrounded by typical capsids (arrowhead), suggesting they may be theresultofaberrantcore

assembly in late infection. Twenty-fourhourspostinfection. Magnification x36,000. 927

F.,)

i

on November 10, 2019 by guest

http://jvi.asm.org/

[image:9.495.55.431.38.534.2]

p

'a

I

.1

S

,

17

J.

-t

A.

,,Nki. - ic~ AO

FIG. 17. An HSV-infected cell showing long stretches of nuclear membrane reduplication as well as

condensation and margination of the chromatin intoseveral large masses. Within the cytoplasm the typical

Golgielements havebeenreplaced by large, apparentlyemptyvacuoles; andnumerousunenvelopedcapsids,

alllackingcoats,arepresent. Tenhpostinfection. Magnification x75,500.

FIG. 18. AnHSV-infected cell in which the typical Golgi membranes have been replaced by concentric

arraysof membrane-boundedcisternae(arrow), andlarge vacuoles.Oftennakedcytoplasmic capsids became

enveloped by buddinginto the latter (inset). Ten h postinfection. Magnification x14,000; inset, x21,000.

928

'$.

.-0 h;0

,:.v., .

N

i;

s

s*S-t Se

v

.. *

t.

_....-r

b'v. i''w'

% ..

*q!\t;*;...<

.<. ...

.. "

w..

$,_

.s @ w

;r2+ R iF

*S .;-

*-r;-.wY . ;

'A

N

..-, I

1,, ..i. .I;illw.

:

x .I ,,

,. e

I ..

.1.

wt

.. k '.

O..:

;#K.

I

V

t

'.-i

"I

I

4

on November 10, 2019 by guest

http://jvi.asm.org/

[image:10.495.66.458.43.620.2]

HSV AND CMV REPLICATION IN WI-38 CELLS

infection when bizarre core forms often appear

(15).

In a system identical toour own, i.e., CMV

strain AD169 growing in WI-38 cells, other

investigators were not able to document capsid

envelopment at the nuclear membrane (11). However, weregularly observed thefullprocess of CMV envelopment at the nucleus, as have

other workers (13).

Despite the observation of envelopment at the nuclear membrane, other capsids of both

HSV and CMV are consistently found unen-veloped in the cytoplasm. Unenunen-veloped

cyto-plasmic CMV capsids were coated, whereas those ofHSVwere not.These results differ from

those in an early report (5) which described

CMV capsids in monkey kidney cells as

un-coated. We found coating tobe a strictly

cyto-plasmicphenomenon, in agreement with the

ob-servation ofStackpole (27) forLuckevirus.

Our studies revealed only one method of nonlytic viral release, namely,fusionbetween a

vacuole carrying an enveloped virus and the

plasmalemma, a process which has been aptly

described as reverse phagocytosis (14). Other investigators (19) have described a tubular

system continuous at one end with the

perinu-clear cisterna, and with theextracellular space at the other, which might serve as a direct

avenueof viral release inHSV-infected HEp-2 cells. Despite numerous attempts to observe such tubules by through sectioning of in situ

embedded monolayers in a variety of selected orientations, we saw no evidence for them in

eitherHSV-orCMV-infected cells.

Wedid observe theformation of

cytoplasmic

densebodiesinCMV-infectedcells inamanner

reported by others (10, 11, 13, 18).

However,

on the basis of

cytochemical

data

reported

else-where(manuscriptin preparation), we feelthat the common assumption that these represent

lysosomesisprobablyerroneous.

With regard to the timing ofthe two cycles, our results agree well with those of previous authors for HSV (15, 24) and with some of the more limited data available for CMV (13, 18).

However, we were not able to identify any

limiting steps in CMV reproduction which

might explain the much longer cycle time. As wehaveshown(manuscript in preparation), the limitation does not appear to involve viral

adsorptionorpenetration, since capsids of HSV and CMV entering the cell from the inoculum both reach the periphery of the nucleus in approximately the same time.

Finally, it should be pointed out that

al-though we have defined several criteria for

discriminating between HSV and CMV

infec-tions in this system, we do not as yet

under-stand the biochemical basis for any of these

distinctions. Especially intriguing isthefailure

of CMV to cause condensation of host cell

chromatin, since this may reflect an effect on host cell macromolecular synthesis which is

different from that of most herpesviruses (12, 17).High-titerCMVinocula, which can now be obtained in a variety of ways (4, 10, 25) should makethis and othersimilarquestionsaccessible

to investigation bybiochemical methods.

ACKNOWLEDGMENTS

We aregratefultoChristopherStackpoleforreading and commentingonthismanuscript.Inaddition,we areindebted toMark Van Boeckelforexcellentphotographicalassistance

andtoDorothy Saltzer for the experttypingofthe

manu-script.

Thesedata were partof aPh.D. thesispresentedby J. D. S. to the Cornell University Graduate School of Medical Sciences.

This investigationwassupported byPublic Health Service Grant CA-08748 from the NationalCancer Institute andby

the Health Research Council of the City of New York,

Contract1-325.

LITERATURE CITED

1. Beale,A.J., and H. C. Hair.1956.Rapiddiagnosisofthe herpetic variety ofKaposi's varicelliform eruptionby

tissueculturemethods. Can. Med. Ass. J.74:443-448. 2. Blank, H., C. F. Burgoon, G. D. Baldridge, P. L.

McCarthy, and F.Urbach.1951.Cytologicalsmearsin diagnosis ofherpessimples, herpeszosterand varicella.

J. Amer.Med. Ass. 146:1410-1412.

3. Brinkley, B.R.,P.Murphy,and L. C. Richardson. 1967.

Procedure forembedding in situ selected cells cultured invitro. J. Cell Biol. 35:279-283.

4. Chambers,R.W.,J. A.Rose,A.S.Rabson,H. E.Bond,

and W.T. Hall. 1971.Propagationandpurificationof

high-titer human cytomegalovirus. Appl. Microbiol. 22:914-918.

5. Dalton,A.I., and R.A.Manaker.1967.Thecomparison of virus particles associated with Burkitt lymphoma

with other herpes-like viruses. In Carcinogenesis: a broad critique, p. 59-90. The Williams& Wilkins Co., Baltimore.

6. deHarven, E., and C. Friend.1965.Originofthe viremia

in murineleukemia. Nat.Cancer Inst.Monogr.no.22, p. 125-134.

7. Doane,F.,A.J.Rhodes, andH.L.Ormsby. 1955.Tissue culturetechniquesinthestudyofherpeticinfections of the eye.Amer.J.Ophthalmol.40:189-193.

8. Flexner, S., and H. L. Amoss. 1925. Contributiontothe

pathologyofexperimentalvirusencephalitis. II. Her-petic strainsofencephalitogenic virus. J. Exp. Med. 41:233-244.

9. Furlong,D., H. Swift, and B. Roizman. 1972.

Arrange-ment ofherpesvirusdeoxyribonucleic acidinthecore. J. Virol. 10:1071-1074.

10. Iwasaki, Y.,T.Furukawa,S.Plotkin, andH.Koprowski.

1973.Ultrastructuralstudyonthe sequenceof human

cytomegalovirusinfection inhumandiploid cells. Arch. GesamteVirusforsch.40:311-324.

11. Kanich, R. E., and J. E. Craighead. 1972. Human cytomegalovirus infection of cultured fibroblasts. II. Viral replicative sequence of awild and an adapted

strain.Lab.Invest. 27:273-282.

12. Kaplan, A. S. 1969. Herpes simplex and pseudorabies

929

VOL.12,1973

on November 10, 2019 by guest

http://jvi.asm.org/

SMITH AND DE HARVEN

viruses. Virol. Monogr. no. 5, Springer-Verlag, New

York.

13. McGavran, M. H., and M. G. Smith. 1965. Ultrastruc-tural, cytochemical andmicrochemicalobservationson

cytomegalovirus (salivary gland virus) infection of

human cells in tissue culture. Exp. Mol. Pathol. 4:1-10.

14. Morgan, C., H. M. Rose, M. Holden, and E.P. Jones. 1959.Electron microscopic observationsonthe

devel-opment of herpes simplex virus. J. Exp. Med.

110:643-656.

15. Nii, S., C. Morgan, and H. M. Rose. 1968. Electron microscopy ofherpes simplex virus. II. Sequence of

development. J. Virol. 2:517-536.

16. Roizman, B. 1971.Herpesviruses, membranes, andthe

socialbehavior of infected cells. p.37-77.In M.

San-ders and M. Schaeffer (ed.), Viruses affecting man

andanimals.W. H.Green, St. Louis.

17. Roizman, B. 1969. The herpesviruses-a biochemical definitionofthegroup.InCurrenttopicsin

microbiol-ogyandimmunology, vol. 49,pp. 1-79. Springer-Ver-lag, New York.

18. Ruebner, B. H., T.Hirano, R. Slusser, J. Osborn, and D. N.Medearis, Jr.1965. Human cytomegalovirus infec-tion.Electron microscopic andhistochemical changes in cultures of human fibroblasts. Amer. J. Pathol. 46:477-496.

19. Schwartz, J., and B. Roizman. 1969. Concerning the

egress of herpes simplex virus from infected cells:

electron and light microscope observations. Virology

38:42-49.

20.Scott, T. F. M. 1964. Herpes simplex. Diagnostic

proce-dures for viruses and rickettsial diseases, 3rded., p.

381-412. InE.H. Lennette andN. J.Schmidt (ed.),

pp. 381-412. American Public Health Ass., Inc., N.Y.

21. Scott,T.F. M. 1965.Theherpes virusgroup.Viral and

rickettsial infections ofman,4thed.,p.892-914.In I.

Tamm and F. L. Horsfall (ed.). J. B. LippincottCo., Philadelphia.

22. Scott, T. F. M., and D. L. McLeod. 1959. Cellular

responses to infection with strainsofherpes simplex virus. Ann. N. Y.Acad. Sci. 81:118-128.

23. Shipkey, F. H., R. A. Erlandson, R. B. Bailey, V. I. Babcock, and C. M. Southam.1967.Virusbiographies. II. Growth of herpes simplexvirus in tissue culture. Exp. Mol. Pathol. 6:39-67.

24. Siminoff, P., and M. G. Menefee. 1966. Normal and 5-bromo-deoxyuridine inhibited development ofherpes

simplex virus. An electronmicroscope study. Exp. Cell Res. 44:241-255.

25. Smith, J. D., and E. de Harven. 1973.Concentrationof herpesviruses. J. Virol. 11:325-328.

26. Smith, K. O., and L. Rasmussen. 1963. Morphology of cytomegalovirus (salivary gland virus). J. Bacteriol. 85:1319-1325.

27. Stackpole, C. W. 1969. Herpes-type virusof thefrog renal adenocarcinoma. I. Virus development intumor

trans-plants maintained at low temperature. J. Virol. 4:75-93.

28. Weller, T. H., J. B.Hanshaw, and D'M.E.Scott. 1960. Serologic differentiation of viruses responsible for

cy-tomegalicinclusion disease.Virology 12:130-132.

29. Wentworth,B. B., and L. French. 1969.Plaqueassayof

Herpesuirus hominisonhumanembryonic fibroblasts.

Proc. Soc. Exp.Biol. Med.131:588-592.

30. Wentworth, B. B., and L. French. 1970.Plaqueassayof cytomegalovirus strains of human origin. Proc. Soc. Exp. Biol. Med. 135:253-258.

31.Wildy, P., W. C.Russell,and R. W. Horne. 1960. The

morphologyofherpesvirus.Virology12:204-222.

32. Wilner, B. I.1969. Aclassificationofthemajorgroupsof

human and other animal viruses, 4th ed. Burgess Publishing Co., Minneapolis.

930

J. VIROL.

on November 10, 2019 by guest

http://jvi.asm.org/