Replication of Sendai Virus

(1)

Replication

of

Sendai

Virus

II. Steps

in

Virus

Assembly

CAROL D. BLAIR' AND WILLIAM S. ROBINSON

VirusLaboratory, University of California, Berkeley, California 94720, and Departmentof Medicine,

Stanford

University,

Stanford,

California

94305

Received forpublication29January 1970

Chick embryo fibroblast culturesinfectedwith Sendai virus wereincubatedwith

3H-uridine in the presence ofactinomycinDbeginningat 18 hrafterinfection.The

35 and 185 virus-specific ribonucleic acid (RNA) components were found in a

ribonuclease-sensitiveforminthecell and appearedtobeassociated with

polyribo-somes. Newlysynthesized 575 viral RNAwasrapidly coated with protein to form

intracellularviral nucleocapsid, and no 57S RNA was found "free"

(ribonuclease-sensitive) in the 2,000 X gsupematant fraction of

disrupted

cells.The

nucleocap-sid from detergent-disrupted Sendai virus and that fromdisrupted cellswere

indis-tinguishable in ultrastructure and buoyant density, and neither was found to be

infectious or have hemagglutinating activity. Kinetic studies ofnucleocapsid and virus formation indicated a relative block in conversion ofviral nucleocapsid to complete enveloped virus inthese cells, resultinginaccumulation oflargeamounts ofnucleocapsid in the cellcytoplasm.

In apreviouspaper (2),weshowedthat several

distinct components of virus-specific ribonucleic

acid (RNA) are synthesized in chick embryo

fibroblast cultures infected with Sendai virus, a

subgroup 2 myxovirus. One component,

repre-senting about one-fourth ofthenewly synthesized

RNA,appears tobeidenticaltothe575 RNA

re-covered from whole virus. The remaining

virus-specific RNA consists of three distinct

com-ponents

sedimenting

at35, 22,and 18S, and 90%

or more of this RNA appears to be

comple-mentary in base sequence to viral RNA, as

indi-cated by its annealing to viral RNA. Previous

studies with a similar subgroup 2 myxovirus,

Newcastle disease virus (NDV), indicated that

virus-specific RNA was synthesized in the cell

cytoplasm

and that much ofthe complementary

RNA was associated with polyribosomes of the

cell (3). The state ofthe 57SviralRNAafter its

synthesis

in cells infected with Sendai virus or

NDV is not known, although some or all of it

may enterthe viralnucleocapsid, as was demon-strated in the cytoplasm of cells infected with

severalsubgroup2myxoviruses (5-7, 10, 13),and

some or all may eventually enter whole virus

whichisformedatthe cell surface (7, 10).

In this study, weattempted to determine the

IPartof this work was taken from a thesis submittedby C. D. B. in partial satisfaction of the requirements for the Doctor of Philosophy degree fromtheUniversityofCalifornia, Berkeley.

locationandstateof the newly synthesizedRNA

components within Sendai virus-infected chick

embryo cells and to study the

incorporation

of

viralRNAinto whole virus. No 57S viral RNA

was found free in the cell cytoplasm in these

experiments. Instead, the 575 RNAappeared to

beimmediately coatedwith

protein

after its

syn-thesis,

making

it

insusceptible

to

digestion

by

ribonuclease.Itwasquantitativelyrecoveredfrom

thecellcytoplasm aspartof the helical viral

nu-cleocapsid. The smaller RNA components were

found in the cell

cytoplasm

in a

ribonuclease-susceptible form, and they sedimented in the

re-gion of polyribosomes and single ribosomes.

Kinetic studies showed that 3H-uridine first

ap-pears in the 57S RNA of the nucleocapsid

re-covered from the cell

cytoplasm;

second, in

nucleocapsid released from cells with trypsin;

third,

in whole virus released from cells with

trypsin; and

finally

inwhole virus in the culture

medium.

Quantitation

of the 575 RNA inthese

pools

indicates thatmorethan95% ofallofthe

newly synthesized 57S RNAremains in the form

of

nucleocapsid

in the cell

cytoplasm,

and that lessthan5%enterswholevirus.

MATERIALS

AND METHODS

Materials. Uridine-5-3H(25c/mmole) and carrier-free H332PO4 were purchased from New England

Nuclear Corp., Boston, Mass. Bovine pancreatic ribonuclease was purchased from Worthington Bio-chemical Corp. Trypsin was purchased from Difco

639

on November 11, 2019 by guest

http://jvi.asm.org/

(2)

and wasprepared as a0.25% solution (w/v) in

Tris-(hydroxymethyl)aminomethane(Tris)-buffered saline. Vibrio choleraeneuraminidase (500 units/ml)was pur-chased fromCalbiochem, Los Angeles, Calif. Tween

80 (polyoxyethylene sorbitan-mono-oleate) was pur-chased from J. T. BakerChemical Co. Actinomycin D was the gift of Merck & Co., Inc., Rahway, N.J. Nonidet 40(NP40)wasthegiftof Shell Chemical Co. Tobacco mosaic virus (TMV) was the gift ofC. A. Knight.

Standard buffer. Thebuffer contained 0.1 M NaCl,

0.01 M Tris-hydrochloride (pH 7.4), and 0.001 M

ethylenediaminetetraacetic acid (EDTA).

Viruses. TheHarris strainof Sendai virus and the

L-Kansas 48 strainofNDV weregrown in

embryo-nated eggs, and the chorioallantoic fluid containing

108 to 109 egg infectious units/mlwas usedto infect

cellsin tissue culture(2).

Cellcultures. Secondary cultures of chick embryo

fibroblasts were cultured and infected with virus as previously described (2).

Assays forinfectivityandhemagglutination. Sendai virusinfectivitywasmeasuredbyanendpointdilution method inembryonatedeggsas previouslydescribed

(2).

Quantitative assay for hemagglutination was done bypreparingserial twofold dilutions of the fluidtobe tested in phosphate-buffered saline (PBS), mixing

with chicken erythrocytes, and incubating as previ-ouslydescribed(2).

Virus purification. Virus was purified from tissue

culturemediumorchorioallantoicfluidbydifferential

centrifugationandsedimentation insucrosegradients

aspreviouslydescribed (2,3).

Isolation of nucleic acids. Nucleic acidfrom cellsor

viruswasisolatedbytheuseof sodiumdodecylsulfate (SDS) and phenol (3). Whenthe amount of nucleic acid in thesampletobeextractedwasverysmall,2mg of TMV carrierwasadded beforephenolextraction.

Preparation ofcytoplasmic extracts. Two methods

wereusedtodisruptchickembryocells. Onemethod has beenpreviously described (3).Cellsgrowingina monolayerwereremovedbyaddition of0.25% tryp-sin and incubatedfor5minatroomtemperaturewith

occasional agitation. Suspended cells were washed withPBS,suspendedinice-coldhypotonicbuffer con-taining 0.01 MNaCl, 0.01 M Tris-hydrochloride (pH

7.4), and 0.002 M MgCl2, and allowedtoswell for 5

min.Cellswerebrokenby 15to 16strokesofa

tight-fitting Dounce homogenizer (Kontes Glass), leaving

most nuclei intact. The broken-cell suspension was centrifugedat4,000rev/minfor 5minin theSorvall

SS34rotorto removeunbrokencells, nuclei,andlarge

membrane fragments. The supernatant fraction was

usedas thecytoplasmicextract. A second methodof

cell disruption used the detergent NP-40 (16). The medium of cell cultures was replaced with 1 to 2

ml of0.5%NP40 in 0.1 M NaCl, 0.01 M Tris-hydro-chloride,and0.002M

MgCl2.

After5min ofincubation withgentle agitation at roomtemperature, cell

frag-mentsremainingonthedishwere scraped off with a

rubber policeman. Nuclei, which remained intact,

andlargecellfragmentswere removed by

centrifuga-tion at 4,000rev/minfor5min, and the supernatant fluidwasusedasthecytoplasmicextract.

Electronmicroscopy. Infected cells were fixed as a monolayer in 1.6% glutaraldehyde in PBS (pH 7.2)

with0.005MMgCl2 and 0.0001 M CaCl2 (PBS MgCa).

Thecells were then scraped off the dish, centrifuged to apellet in McNaught protein tubes, and fixed as a pellet in1% osmic acid in PBS MgCa. After dehydra-tion in agraded ethylalcohol-0.15 M NaCl series, in-filtration with propylene dioxide, and embedding in Epon 812, thin sections were cut on an LKB ultra-microtome. Contrast was increased by staining with 1% aqueous uranyl acetate and 1% alkaline lead citrate (14).

Virus or nucleocapsid was pelleted in the Spinco SW50 rotorby spinning at 47,000rev/min for 90min. Thepellet was resuspended in a small volume of dis-tilled water, and a small drop of this suspension was delivered to a collodion-covered 200-mesh copper grid. Negative stain was2%ophosphotungstic acid

ad-justedto pH7with KOH.

Micrographs were taken on a SiemensElmiskop I electron microscope at either 60 or 80 kv with a

50-Mzm

objectiveaperture.

Scintillation counting. Radioactivity precipitated in 5% trichloroacetic acidwascollected and washed on Millipore filters and counted in toluene scintillation fluid inaPackardTri-Carb scintillation spectrometer

(3).

RESULTS AND DISCUSSION

Fractionation of cytoplasmic extracts from Sendai virus-infected cells. Chick embryo fibro-blast cultures were incubated with 3H-uridine in

the presence ofactinomycinD from 18 to 20 hr after infection with Sendai virus, when virus-specific RNA synthesis is maximum (2). Under theseconditions,alllabeledRNAhas beenshown

to bevirus-specific RNA (2). A cytoplasmic ex-tract wasthenprepared byusing NP40 detergent and divided into three partsforsedimentation on

sucrosedensity gradients.Figure IA shows the dis-tribution in a sucrosegradient of trichloroacetic acid-precipitable radioactivity and cytoplasmic

materialabsorbing at 260nmaftersedimentation of a sample of the cytoplasmic extract layered

directly ona sucrose gradient. TheA260 peak in

fraction 23 representssingleribosomes(74S), and that in fraction 27, the 50S ribosomal subunit.

The ultraviolet-absorbing material sedimenting

morerapidly than single ribosomes (fractions1to 20) represents polyribosomes. A significant frac-tion of theradioactiveRNAsediments in a broad distribution with polyribosomes and the single

ribosomes.Adistinct radioactive component with

apeakinfraction 8 sediments morerapidlythan

the single ribosomes and most of the

polyribo-somes detected by ultraviolet absorbancy. Some

radioactive material sediments more slowly than

singleribosomes.

Figure1Bshowstheresults of sedimentation of

asecondsampleofcytoplasmicextractafter

incu-J. VIROL.

on November 11, 2019 by guest

http://jvi.asm.org/

(3)

SENDAI VIRUS REPLICATION

Fract,on Number

FIG. 1. Sedimentation of cell extractsfrom Sendai virus-infected cells. Four 150-mm tissue culture dishes

with2 X 107 cells eachwereinfected withSendai virus. Eighteenhours later, the culture mediumof each was

re-placed with 10mlof medium containing 2 ,ugofactinomycin D perml; 30 min later,300 ,uc of3H-uridinewas

addedtoeach culture. After2hr, theculturemedium was removed, and a cytoplasmic extract was preparedfrom the cellsofallfourculturesby using0.5%NP40. The 4,000 rev/minsupernatantfractionwasdividedintothree

parts. Thefirstportion (A) waslayered directlyonto agradientof5 to

30'7j

sucrose in0.1MNaCI, 0.01 M Tris

(pH 7.4), and0.002 MMgCl2. Thesecond portion (B) was incubated with 5 ,ug of pancreatic ribonuclease for 5

min at roomtemperature beforelayeringon asecond sucrosegradient. ED TA wasadded to a third portion (C)

tomakeafinalconcentrationof0.004 Mbefore layeringthe extracton athird sucrosegradient.All three sucrose

gradients werecentrifuged for2.5 hr at 25,000rev/minand 4 C in aSpincoSW 25.3 rotorand analyzedfor

tri-chloroacetic acid-precipitableradioactivity andA260aspreviously described (2).

bation withpancreatic ribonuclease. The increase

in

ultraviolet-absorbing

material in the

single-ribosome peak (around fraction 23) and the

de-crease in polyribosomes, indicated by the

de-creaseinabsorbancy around fraction15 (compare

Fig. 1A), resulted from ribonuclease action on

polyribosomes. It can be seen that ribonuclease

did not alter the sedimentation behavior of the

rapidly sedimenting radioactive component

around fraction8ortheamountoftrichloroacetic

acid-precipitable radioactivity

in this component.

Figure 1C shows the results of sedimentation

ofathird

portion

of

cytoplasmic

extracttowhich

EDTA had been added. The reduction in the

amount of

ultraviolet-absorbing

material in

frac-tions 1 to 23 is consistent with dissociation by

EDTA ofpolyribosomes into more slowly

sedi-menting components. Italso appears thatEDTA

did not alter the sedimentation behavior or the

amountof trichloroaceticacid-precipitable

radio-activity in the rapidly sedimenting component

around fractions 9 and 10. The amount of this

RNA-containing component appears to be too

smalltobedetectedbyabsorbancyat260nm.

These results suggest that some newly

synthe-sized RNA in Sendai virus-infected cells

sedi-mentswithpolyribosomes, single ribosomes, and

a significant fraction ofnew RNA in a rapidly

sedimenting

structure. The sedimentation

be-havior of this structure isnotaltered by

ribonu-cleaseorEDTA, andthelabeledRNAassociated

withitisnotmadetrichloroacetic acid-soluble by

ribonuclease. In experiments in which infected

cells were disrupted by a Dounce homogenizer

without NP40detergent, aspreviously described,

the sucrose gradient profiles and the amount of

radioactiveRNAineach component inthe

gradi-ents werethesame asin theexperiment shown in

Fig. 1, indicating that the rapidly sedimenting

structure doesnotresult from theaction ofNP40

on infected cells or whole virus associated with

the cells.

Identification of RNA in cytoplasmic extracts.

To identify the RNA in different regions of the

sucrosegradient shown in Fig. 1A, the fractions

in each of the regions designated 1 to 4 were

pooled, and the RNA was recovered from each

byphenolextractionbyusing50,ugofTMVRNA

as carrier. The purified RNA from each region

was then fractionated by sucrose gradient

sedi-mentation (Fig. 2).

Figure 2A shows the results with the RNA fromregion1 inFig.1A.Sedimentationwasfrom

rightto left. The component of RNA detectedby

641

VOL. 5, 1970

on November 11, 2019 by guest

http://jvi.asm.org/

[image:3.486.48.435.67.262.2]

(4)

Prac*on Niuber

FIG. 2. Sedimentation of RNA extractedfrom fourregionsof thesucrosegradient showninFig. IA.All

frac-tions ineachof the numbered regions of thesucrosegradientinFig. JAwerepooled,2mgofTMVwasaddedas

carrier, andRNAwas extractedfrom each. Each RNA sample wasthensedimentedin asucrose densitygradienlt

in aSpinco SW 50rotor at47,000 rev/min and4Cfor105 min aspreviously described (2).

A260 with a peak in fraction 18 (Fig. 2A) is the

small amountofTMV RNAused ascarrier.No

distinct component of radioactive RNA is

ap-parent.

The RNAfromregion2 inFig. IA, which

in-cludes the rapidly sedimenting radioactive

struc-ture, is shown inFig. 2B. The RNA component

detected byA260with a peak infraction 15 (Fig.

2B) representsTMV RNAandasmall amountof

28SribosomalRNAwhich sediments withTMV

RNA under these conditions. Almost all of the

radioactiveRNAsediments like 57SviralRNA,

and no significantamount ofradioactive 18, 22,

or35Svirus-specificRNAis present.

Theresults ofsedimentation of the RNAfrom

region 3 in Fig. 1A, which includesmost ofthe

polyribosomes and single ribosomes detected by

ultraviolet absorbancy,areshowninFig.2C.The

RNA componentsdetectedby A260,withpeaks in

fractions 14and 19 (Fig. 2C), consist mostly of

28and18Sribosomal RNA,respectively. Twoof

the components of radioactive RNA sediment like the355(fraction 9)and the 18S (fraction 19)

virus-specific RNA forms, which have been

showntobealmostcompletely complementaryin base sequence to viral RNA (2). Some

radioac-tive RNA (fraction 25) sediments more slowly than 18S and may represent virus-specific RNA

degraded duringthe preparationof the

cytoplas-mic extract. No labeled57S RNA was recovered

from this region of the gradient (e.g., region 3,

Fig. 1A) in severalexperiments.

The RNAfromregion 4inFig. 1A, which

in-cludes material sedimenting more slowly than

single ribosomes, is shown in Fig. 2D. Again,

ribosomal RNA is detected by A260, and almost

all of the radioactive RNA (fraction 25)

sedi-mentsmoreslowlythan 18S.

Theseresults suggestthatall newly synthesized

57S virus-specific RNA from Sendai

virus-in-fected cells can be recovered from the rapidly

sedimenting ribonuclease-resistant structure (Fig.

1), and the35and 18Svirus-specificRNAforms

sediment in a ribonuclease-sensitive form with

polyribosomes and single ribosomes. The latter

finding is in agreement with experiments with

NDV-infectedcells, in whichthe 18 and 35S

virus-specific RNA components sedimented with

polyribosomes (3).

Identification of the rapidly sedimenting

ribo-nuclease-resistant structure in cytoplasmic

ex-tracts as viral nucleocapsid. A number of the

properties of the

rapidly

sedimenting

ribo-nuclease-resistant structure from infected cells

(Fig. 1) indicate that it is a viral nucleocapsid.

Consistent with this arethe observationsthatits

sedimentation was not altered by ribonuclease

(Fig. 1B)orEDTA(Fig.IC),thattheradioactive

RNAassociatedwiththe structure was not made

trichloroacetic acid-soluble byribonuclease (Fig.

1B),andthatRNAwith sedimentation character-istics of viral RNA (57S) was exclusively

re-covered from this component in many

experi-ments. Kingsbury and Darlington (12) have shown that the nucleocapsid recovered from

de-tergent-disrupted NDV contains high molecular

weight RNA,andthat ribonucleasedoesnotalter

thesedimentationrateof the nucleocapsidor de-grade the RNA to a trichloroacetic acid-soluble

form.

on November 11, 2019 by guest

http://jvi.asm.org/

[image:4.486.65.450.63.228.2]

(5)

The buoyant density of the rapidly sedimenting

structure from Sendai virus-infected cells,

de-termined by equilibrium centrifugation in a

pre-formedsucrose density gradient, was 1.26to 1.27 g/ml. (Fig. 3). This value is significantly greater than the buoyant density value of whole virus, which is 1.22 g/ml under the same conditions.

Compans and Choppin (5) found a buoyant

density of1.30g/mlinaCsCldensity gradientfor

the nucleocapsid isolated from SV-5-infected BHKcells.

Figure 4 shows the results of rate zonal sedi-mentation of the rapidly sedimenting component

(closed circles) in a 5 to 20% sucrose density gradient. TMV (opentriangles),which hasa

sedi-mentation coefficient of about 160 S20,w under these conditions, was used as a sedimentation

marker in the experiment. The radioactive

cyto-plasmic component clearly sediments more

rapidly than TMV and can be estimated to

haveasedimentation coefficient ofapproximately 200S under these conditions. The sedimentation

coefficient of intact NDV, a virus with almost identical size andshapeas Sendaivirus,hasbeen estimated to be 1,100S (15). Thus, whole virus

has a much greater rate of sedimentation than that of the cytoplasmic component in Fig. 4.

3,000

, 2,000

0

E

H9-1,000

5 10 15

f raction number

20

FIG. 3. Equilibrium centrifugation of viral

nucleo-capsid. 3H-labelednucleocapsid, preparedasdescribed

inthelegendtoFig.IC,waslayeredoverapreformed

15 to65% sucrosegradient containingstandardbuffer

andcentrifugedat47,000 rev/minl for4 hrat4 C ina SpinicoSW50rotor.

0

09 C\J

0-U

l<

Fraction Number

FIG.4. Velocity sedimentation of nucleocapsid.

3H-labeled nucleocapsid,preparedas describedinFig.

IC, wasmixedwith unilabeledTMV, anzdthe mixture

waslayeredovera5 to 20% sucrosedensity gradient

andcentrifugedat45,000rev/minzand 4 C for 15 min inanS W50rotor.

Kingsbury and Darlington (12) have estimated

the sedimentation coefficient of the nucleocapsid from detergent-disruptedNDVtobe200S.

Todetermine themorphological appearance of

the200S cytoplasmic structure, it was examined

by electron microscopy after negative staining.

Whole virus grown in the chorioallantoic cavity

was also examined for comparison. Figure 5A

shows that typical viral nucleocapsid is present

in thesucrosegradient in the position of the 200S

structure. The nucleocapsid is indistinguishable

from that described in detail from SV-5-(5) and NDV-(6) infectedcellsandfrom disrupted Sendai

virus (4, 9) and NDV (12).

Figure SB shows whole Sendai virus. The

nu-cleocapsid within virions and that lying outside

have the same morphology as the cytoplasmic cleocapsidinFig. 5A.

Comparison of cytoplasmic nucleocapsid with

wholevirus anddetergent-disruptedvirus. Further

evidence that therapidly sedimenting cytoplasmic

structure is viral nucleocapsid was obtained by

testing its infectivity and hemagglutinating

ac-tivity (HA) andby comparingit with whole virus and detergent-disrupted virus. Figure 6A shows theresults ofequilibrium sedimentation Of32PO4_ labeled Sendai virus after disruption with 0.3%

VOL. 5, 1970

643 on November 11, 2019 by guest

http://jvi.asm.org/

[image:5.486.242.432.65.304.2] [image:5.486.39.229.347.582.2]

(6)

sodium deoxycholate (DOC) and layering over a preformed sucrosedensity gradient. The heteroge-neous32P-labeled component with apeakin

frac-tion6has the buoyantdensity of the nucleocapsid

isolated from cells (1.27 g/ml). It was demon-strated that almost all of the 32p in this

com-ponent afterphenol extractionwas in theform of

57S RNA,which is consistent with its being viral

FIG. 5. Morphology ofcellularnucleocapsidand virus. (A) Nucleocapsidwaspreparedfrom infectedcells as

describedinFig. 1,andsucrosegradientfractionscontaining 3H-nucleocapsidwerecentrifuged to make a pellet

forelectron microscopyasdescribedinthetext. X240,000. (B) Sendai virus growninembryontatedeggs

andpuri-fiedin sucrosegradientswaspreparedforelectronmicroscopyasdescribedinthe text. X 100,000.

on November 11, 2019 by guest

http://jvi.asm.org/

[image:6.486.93.425.113.602.2]

(7)

Fraction Number

FIG. 6. Comparison of nucleocapsidfrom cells, nucleocapsid from virus, and intact virus. Sendai virus-labeled with32PO4 waspreparedinembryonatedeggs andpurified as previously described (2). To one sample Of 32pO4 virus (A), DOCwasaddedto afinalconcentrationof0.3%,and the mixture was shaken for5min at room tem-peratureandthenlayeredover a15to65% sucrosegradient.Asecondsampleof 32PO4 virus (B) was layered di-rectlyon an identicalsucrosegradient. A cytoplasmic extract was preparedfrom seveninifected150-mm cultures after incubationz with 3H-uridineandactinomycinD asdescribedinFig.1, and the extract (C) was layered over a third 15to65%' sucrose densitygradient.

All threesampleswerecentrifugedfor 14 hrat25,000rev/min and 4 CinaSpinco SW 25.3 rotor. Fractions werecollected dropwisefrom the bottom of the tube, and trichloroacetic acid-precipitable radioactivity was

deter-minedfor 0.1-ml samples. A 0.05-mlamount wasremovedfrom each fraction, diluted in serum-free 199A, and

frozenfor laterinfectivity and HA assays as describedin the text. Densities were determined by weighing

100-pliterportionsinamicropipette.

nucleocapsid released from DOC-treated virus.

No viral infectivity was associated with

DOC-treated virus, and HA was only associated with

material sedimentingnearthetopofthe gradient

(fractions 10 to21). Figure6B showstheresults

of sedimentation of 32P-labeled whole virus before

treatment with DOC and demonstrates that

radioactivity, infectivity, and HA coincide at a

density

of 1.23

g/ml (around

fraction12).

Figure

6C shows the results of equilibrium

sedimenta-tion ofa cytoplasmic extract from infected cells

prepared asdescribedfor the experiment in Fig.

1. Thetritium-labeled componentwith apeak in

fraction6has thebuoyantdensityofnucleocapsid

(1.27 g/ml). The material with HA has a peak

witha

density

around1.22to1.23g/mland

repre-sentswholevirus. Infectivity andHA do not

ap-peartobespecificallyassociated with the

nucleo-capsid in this experiment, and purified

nucleo-capsid in otherexperiments failed to show

infec-tivityorHA.

These

experiments

suggest that nucleocapsids

from detergent-disrupted Sendai virus and from

the cytoplasm of infected cells are similar in

buoyant density, contain 57S RNA, and do not

have infectivity or HA. Although Bukrinskaya

etal. (4) havereportedinfectivity associated with

nucleocapsid from Sendai virus-infectedcells, we

have beenunable to demonstrateinfectivity with

ournucleocapsid preparations.

Formation of nucleocapsid. To determine the

time required for nucleocapsid assembly

after viral RNA synthesis, an experiment was

done in which the rate of 3H-uridine

incorpora-tion into total 575 RNA and into intracellular

nucleocapsid was measured, starting at 18 hr

afterinfection.Different cultures ofinfected cells

wereincubated with high specific activity

3H-uri-dinefor 5, 10,and 15 min inthepresenceof

acti-nomycinD, and theamounts ofradioactivity in

total57S RNA ofthecells and theradioactivity

inintracellular nucleocapsidweredetermined for

each incubation time. It was shown earlier (Fig.

2B) that all of the radioactive RNA extracted

fromthenucleocapsid is57SRNA.

The results in Table 1 show that, after only 5

min ofincubation with

3H-uridine,

all of the la-beled 57S RNAin the cell cytoplasm canbe

ac-counted for by radioactivity in nucleocapsid.

Similarly, after 10 and 15 minofincubation with

3H-uridine, essentially all of the radioactivity in

57S RNAcanbe accounted for in

nucleocapsid.

The differences between theamounts of

radioac-tivity in the two fractions at all three times are

10% or less, a value which is within the

experi-mental error of the methods used. Two conclu-645

VOL. 5, 1970

on November 11, 2019 by guest

http://jvi.asm.org/

[image:7.486.44.439.67.249.2]

(8)

BLAIR AND ROBINSON

TABLE 1. Incorporation of 3H-uridine into total 57S RNA in infected cells and into viral

nucleocapsida

Counts/min of BHlabel in Time of 3H-uridine

incubation

57S RNA Nucleocapsid

min

5 822 856

10 1,828 1,724

15 8,548 7,718

a

Six

150-mm culture dishes containing2 X 107

cells each wereinfectedwith Sendai virus.

Eight-een hours later, the medium of each culture was replaced with 10 ml of medium containing 2ug of actinomycin D per ml. Thirty minutes later,

400 scof3H-uridinewasadded to eachand,atthe subsequent times designated, two culture dishes

wereplaced inan ice-waterbath and themedium

replaced with ice-cold phosphate-buffered solu-tion.Cells wereremoved from thedishes with con-centratedtrypsinin thecold, and the cell

suspen-sionswere divided intotwoparts.From onepart,

RNA wasextractedwith sodium dodecyl sulfate andphenol, and theradioactivityin575RNA was

determined after sucrose gradient sedimentation

of the RNA (2). The second part of each cell

suspension was disrupted with NP40, and the

radioactivity in nucleocapsid was determined

after sucrose gradient sedimentationasdescribed forFig.1.

sions can bemade from this experiment. First,

57S RNAbecomes partoftheviralnucleocapsid

very rapidly after its synthesis. The time is

un-doubtedly significantly

less than 5 min, since

es-sentially all57S RNAis found innucleocapsidat

5 min. This rapid assembly of Sendai virus

nu-cleocapsid is similar to the maturation ofpolio

virus, which has been estimatedto occurwithin3

min orless of thecompletion of synthesisofviral

RNA (1). Second, not only is57S RNA rapidly converted tonucleocapsid after its synthesis, but

all or avery

large

fraction of it entersviral

nu-cleocapsid. Free57S RNAisnotdetected byour

methods in cell cytoplasm at this stage of the

virus growth cycle. The results described inFig.

1A and 2 arein accordwiththeseconclusions.

The resultshere suggest that thesmallerRNA

componentswhich appear to be associated with

polyribosomes and not the 57SRNAmay serve

asviralmessenger. It is notexcluded thatthe 57S

RNAdoes enterpolyribosomesat an

earlier

stage

in the virus

replication

cycle. Ithasbeen shown

thatmost of

newly

synthesized poliovirus RNA becomes associated with polyribosomes early in

thepolio growth cycle, andmost enterscomplete

virions latein thecycle (1).

Virus formation at the cell surface. The final

steps inmyxovirus assembly are thoughtto take place atthe cell surface by theprocessdescribed

asbudding (7, 11).We havefound thattreatment

of Sendai virus-infected cells with trypsin quanti-tativelyremovescellassociated virus without

dis-rupting cells. This was shown by an experiment

in which cells, after incubation with 8H-uridine

in the presenceof2,ug of actinomycin D per ml

between 18 and 20 hr after infection with Sendai virus, were gently removed from culture dishes

withtrypsin solution and centrifuged, and the

re-sulting supernatant fluid was analyzed by

equi-librium centrifugation in a preformed sucrose

density gradient. Figure 7 shows the distribution in the gradient of the radioactive particulate

ma-terial released from the infected cellsby trypsin.

Two incompletely separated radioactive

com-ponentsappearin the gradient. One has the

buoy-antdensity of whole virus (1.23 g/ml) and is

asso-ciated with infectivity and HA. The other

L 130 Q 0) 0N26 U) c122 °~1,14 200~ E 0I to 100 104_. L_ a) D _-*£ 14 Vo E

0.

120I 40

24 6 8 10 12 14 16 Fraction N\umber

FIG. 7. Particulate material releasedfrom cellsby

trypsin. Seven 150-mm culturesofSendaivirus-infected

cellswereincubated with 100,c of3H-uridineper

cul-tureand actinomycin D asdescribed in thelegendto

Fig. 1. The culture mediumwasthenremoved,thecells

werewashedwithPBSwithout CaorMg,and1 mlof

0.25%trypsin solution wasaddedtoeachfor5 minof

incubationatroom temperaturewithoccasionalgentle

agitation. The trypsinsolutionandfloatingcellswere

removed,and each dishwaswashedwithImlof PBS.

Thepooledtrypsin and washsolutionswerecentrifuged at4,000rev/minfor5 mintoremovewhole cells; the supernatant fraction was layered over two sucrose layers (Imlof65% sucroseinD20and 5 mlof15% sucrose) in acentrifuge tube andcentrifuged for2hr at25,000rev/min and 4 C in theSpincoS W25.3rotor as previously described for viruspurification (3). All

oftheparticulate material which included whole virus

andnucleocapsidatthe interface between the sucrose layers wasrecovered andlayeredover a linear 15 to

65%sucrosedensity gradientandcentrifugedfor12hr

at25,000rev/min and 4 CinaSpincoS W 25.3rotor.

646

J. VIROL.

on November 11, 2019 by guest

http://jvi.asm.org/

[image:8.486.52.241.66.185.2]

(9)

component has the buoyant density of viral

nucleocapsid, and electron microscopyofmaterial

from this region of the gradient reveals typical

structures of viralnucleocapsid.

When 3H-RNA was isolated from each com-ponent, it was found to sediment as 57S viral RNA. Thus, it appears that the action of trypsin

resulted in release ofboth infectious virus and

viral nucleocapsid from infected cells. After

incu-bation ofinfected cells with H3-uridine for 2 hr,

significantly more radioactivity appeared in the

RNAof thetrypsin-released nucleocapsid than in

trypsin-released wholevirus.

Figure 8 shows a comparison of the effects of

trypsin, neuraminidase, and EDTA on release of

virusandnucleocapsidfrominfectedcells.

Identi-calcultures ofinfectedcells afterincubationwith

3H-uridine in the presence ofactinomycin D, as

described forFig.7, were treatedwithtrypsin

so-lution (Fig. 8A), V. cholerae neuraminidase at a

concentration 10timesthatrequiredtoinhibit

ag-glutination of an equivalent number of chicken erythrocytes (Fig.8B), or a versenatesolutionat a

concentrationwhich removesepithelial cellsfrom

culture dishes (Fig. 8C). The particulate material

released by trypsin canagainbe seen to be

frac-tionated by sucrose gradient centrifugation into

twocomponentscorresponding towhole virus at

a density of 1.23 g/ml and nucleocapsid at a

density

of 1.27 g/ml (Fig. 8A). Very small

amountsof whole virusandalmost nodetectable

nucleocapsid were released from cells by

neura-minidase

(Fig. 8B)

or EDTA

(Fig. 8C),

indicat-ing that the whole virus and nucleocapsidreleased

withtrypsinarenotbound to cellsby bonds

dis-ruptedby neuraminidase or EDTA.

Incubation of purified whole 3H-virus with trypsin at room temperature and subsequent frac-tionation ofthe 3H-labeled virus preparation by

sucrose gradient sedimentation, as described for Fig. 7, demonstrated that nonucleocapsid was re-leased from whole virus by trypsin alone. This in-dicates that nucleocapsid is released from cells rather than whole virus after incubation of in-fected cells with trypsin. Whether the nucleocap-sid in these experiments comes from a few cells which are mechanically disrupted during the

trypsin treatment and centrifugation, or whether

trypsin more selectively releases nucleocapsid

from sites near or at the surface of infected cells, is not completely clear. An experiment to test

these possibilities suggests the latter. Cells were

incubated with 3H-uridine between 11 and 25 hr afterinfection in the absence of actinomycin D to

permit cellular RNA, such as ribosomal RNA

and virus-specific RNA, to become radioactive.

The cells were then gently removed from the dishes with trypsin and collected by centrifuga-tion. The totalradioactive RNA in the cells was determined, a sample of the trypsin solution was used to determine totalradioactivity released by trypsin, and a second sample was used to

deter-mine theradioactivityin the RNA in

nucleocap-sid and whole virus which were isolated from the

E

0I

Fraction Number

FIG. 8. Particulate material releasedfrom cells by trypsin, neuraminidase, andED TA. Six 100-mm cultures

with 6 X 106cellseach wereincubatedwith 100 xsc of3H-uridineperculture andactinomycin Dbetween 18 and

20hrafterinfection, asdescribedinthelegendtoFig.1. (A) The cells oftwocultureswereincubatedwithtrypsin

andwashedasdescribed in thelegendtoFig.7. (B)Fiftyunitsof Vibrio cholerae neuraminidaseinImlwasadded

to thesecondpairof cultures, whichwere then incubatedat room temperaturefor20 min. Neuraminidase and

floating cells were removed, and I mlofPBS was used to washeachplateasinpart A. (C)A

1-ml

amountof

0.001M sodium versenate in PBS without Ca2+ andMg'+ wasaddedtothe thirdpairofcultures. Thecultures

were incubatedfor20 min at room temperature, and the disheswere washedasforparts Aand B. The

super-natant fraction ofeachsample, whichwascentrifugedat4,000rev/mintoremovewholecells,wasthen centrifuged

insucrosegradientsasdescribedinthelegendtoFig. 7.

VOL.5, 1970 647

on November 11, 2019 by guest

http://jvi.asm.org/

[image:9.486.103.396.400.548.2]

(10)

AND

trypsin solution in a sucrose density

gradient,

as

inFig. 7. Table 2 shows that about 1.5% ofthe total radioactive RNA of the cells was released into the trypsin solution, and, of the

trypsin-re-leased material, about 20%wasRNA in nucleo-capsid. We have also shown that the amount of radioactive nucleocapsid and virus released by trypsin is 15 to 20% of the amount of

intra-cellular radioactive

nucleocapsid

(Fig. 9). Thus,

amuch greaterproportionof totalcell-associated

radioactive

nucleocapsid

than total

radioactive-cellRNA, mostof which is ribosomal and trans-fer RNA, is released by trypsin. This indicates that trypsin

preferentially

or

selectively

frees

nucleocapsidfrom infectedcellsandsuggeststhat

nucleocapsid may accumulate at or near the cell

surface. In the case of SV5-infectedmonkey

kid-neycells,viralnucleocapsidhas been shown to

ac-cumulate immediately under some areas of the cellsurface (7).

Kinetics ofvirus

assembly.

We have described four structures containing 57S RNA which can

be recovered from Sendai-infected cells or cell culture medium: intracellular nucleocapsid,

nucleocapsid, virus released from cells by

trypsin, and virus in the culture medium. An

experiment was then done to determine the time course of 3H-uridine

incorporation

into theRNA of each ofthese structures.

Actinomycin

D-treated cells were incubated with 3H-uridine from 18 to

22 hr after

infection,

and the four particulate componentsnamed above were isolated atvarious

times throughout this period. The cells were

removed withtrypsin asdescribedfor

Fig. 7,

and

TABLE 2. Release of tritium-labeled RNA from

infected cellsbytrypsina

Total acid-Fraction precipitable 8H-uridine (counts/min)

Total cell RNA

.4,204,000

Total RNA released by trypsin 64,210 Nucleocapsid RNA released by

trypsin.13,428

aFour 100-mm cultures were each incubated

with 100,c of3H-uridine in 5 ml of medium

with-outactinomycinDfrom11 to 25hr afterinfection

with Sendai virus. Cellswere then removed with

trypsin and collected by centrifugation as de-scribed in Fig. 7. The total RNA was extracted from cells and from one-half of the trypsin so-lution by the sodium dodecyl sulfate-phenol method; a sample was used for trichloroacetic

acidprecipitationandcounting.The other half of the trypsin solution was centrifuged in sucrose

density gradients for isolation of nucleocapsid

andvirus,asdescribed in thelegendtoFig. 7.

E 14,000

-10,000

-6,000

-2,000- T

20 40 60 80 100 120 140 60 180 200 220 240

Time,minutes

FIG.9. Incorporation of3H-uridine intofour viral

structurescontaining 57S RNA. Five milliliters of

me-dium with2 ,ug ofactinomycinDperml was addedto

each

offive

100-mmculturesat17.5hrafterinfection

withSendai virus; 30 min later, 100 Muc of3H-uridine

was added to each culture, and incubation was

con-tinuedat37C. Theincubation of individual cultureswas

stopped5,30, 60,120,and 240minlater by placing the

dishinanicebath andreplacingthemedium with

ice-cold PBS.Cellswereremovedwith0.25%trypsin in the

cold, collected by centrifugation, and disrupted by

Dounce homogenization as described in the text.The culturemedium, trypsin solution, and cytoplasmic

ex-tract for each incubation time were cenztrifuged

in sucrose density gradients to isolate niucleocapsid

and intact virus. The sum of the trichloroacetic

acid-precipitable

radioactivityin thenucleocapsidfrom

thecytoplasmic extract or homogenate (H),

nucleo-capsidfrom

thetrypsinsolution(T,),intact virusfrom thetrypsinsolution (T2),andintactvirusfrom the

cul-turemedium

(V)

areplottedforeachincubationtime.

thencellswere

disrupted

by

a Dounce

homoge-nizer without NP40. Virus in the medium and

whole virusand

nucleocapsid

released by trypsin

at each time were separated (Fig. 7), and the

nucleocapsid

was recovered from the

disrupted

cells

(Fig. IC).

Figure

9showsthe resultsofthis

experiment,

and it can be seen that 3H-uridine

firstappearsincytoplasmic nucleocapsid; second,

in

nucleocapsid

releasedbytrypsin; third, invirus

released

by

trypsin;

and finally in virus in the

culturemedium. Thisorderiscompatible withthe

structures

being

formed in a stepwise sequence

during

virus assembly.In addition, during the

ex-periment, intracellular

nucleocapsidprogressively

accumulates,

since the rate of its formation is

significantly

greaterthan itsrate ofconversionto

virus.

Serial electron micrographs ofSendai-infected

cells

(Fig.

10) document

this accumulation. By 18 hrpostinfection (Fig. 10A),largeaggregatesof

viral

nucleocapsid

are evident in infected cell

cytoplasm. By

4

days

after infection (Fig. lOB), the size of

nucleocapsid

aggregates has

enlarged

J. VIROL.

on November 11, 2019 by guest

http://jvi.asm.org/

[image:10.486.256.447.69.212.2] [image:10.486.48.239.443.525.2]

(11)

FIG. 10. Electronmicroscopy of Sendaivirus-infectedchick embryofibroblasts. Cellsat 18 hr(A) and4days (B)werepreparedfor electron microscopy as describedin the text.Magnlification:A, X 54,000; B,44,000.

VOL. 5, 1970 649

on November 11, 2019 by guest

http://jvi.asm.org/

[image:11.486.58.417.66.585.2]

(12)

BLAIR AND ROBINSON

until much of the normal cytoplasmisdisplaced.

In other subgroup 2 myxovirus-infected cells, such as SV5-infected BHK cells (7) and

para-influenzatype2-infectedFLand HeLa cells (10), similar accumulation of viral nucleocapsid in

infected cell cytoplasm has beenshown.

These results indicate that, relative to the rate

ofviral RNA synthesis and its rapid conversion

to nucleocapsid in Sendai virus-infected chick

embryo

fibroblasts,

conversion of nucleocapsid

to matureenvelopedvirus isaveryslowor

ineffi-cient step in virus formation. The radioactivity

data suggest that only approximately 5% of

newly synthesized viral RNA leaves the cell in

whole virus and the remainder progressively accumulatesinthe cellintheform ofviral

nucleo-capsid. The low yield of infectious virus by

cul-turesinwhichalmost allcellsareinfected (2) and

the serial electron micrographs showing large

amounts of nucleocapsid within infected cell

cytoplasm are in accord with this. With SV-5, another subgroup 2 myxovirus, there isevidence

that thisbehavioriscell-specific (7, 8).Thereason

for the relative block in virus formation is not

clear, but it could be due to a limiting rate of

synthesis ofanessential component of the virus,

such as a viral envelope protein. Experiments

areinprogresstocharacterize thevirionproteins

and attempts will be made to learn whether any

virus component is synthesized in limiting

amounts.

ACKNOWLEDGMENTS

Thisinvestigationwassupportedby Public HealthService re-searchgrants CA 08557andCA 10467from theNationalCancer Institute andtraininggrantTIGM 1389from theNational Insti-tuteof GeneralMedicine.

We thank Jeana Levinthal, Robley C. Williams, Charlotte Lizarraga,andJoseph Toby for assistance with electron

micros-copy.

LITERATURE CITED

1. Baltimore,D.,M.Girard,andJ. E.Darnell. 1966.Aspectsof thesynthesisofpoliovirusRNA and theformationofvirus particles. Virology 29:179-189.

2.Blair, C. D.,andW. S.Robinson.1968.ReplicationofSendai

virus.I.Comparisonof viral RNA andvirus-specific RNA

synthesistoNewcastle disease virus. Virology 35:537. 3. Bratt, M. A., andW. S. Robinson. 1967. Ribonucleic acid

synthesisin cellsinfected withNewcastlediseasevirus. J. Mol. Biol. 23:1-21.

4.Bukrinskaya, A. G.,S.M.Klimenlco,J.A.Smirnov,and B. V. Guschin. 1968. Infective substructures of Sendai virusfrom infectedEhrlich ascitestumorcells. J. Virol. 2:752-758. 5.Compans,R.W.,andP. W. Choppin. 1967a. Isolationand

propertiesof thehelicalnucleocapsidof theparainfluenza virusSV5.Proc. Nat. Acad. Sci. U.S.A.57:949-956. 6.Compans,R.W.,and P.W. Choppin. 1967b. The lengthof

the helicalnucleocapsidofNDV.Virology 33:344-346. 7.Compans,R.W.,K.V.Holmes,S.Dasel,andP.W.Choppin.

1966. An electron microscopic study of moderate and virulent virus-cell interactions of the parainfluenza virus SV5. Virology 30:411-426.

8.Holmes, K. V., and P.W.Choppin. 1966. On the role ofthe

response of the cellmembrane indetermining virus viru-lence;contrasting effect of the parainfluenza virus SV5 in

twocelltypes.J.Exp. Med.124:501-519.

9. Hosaka,Y.,H.Kitano,andS.Ikeguchi. 1966.Studiesonthe

pleomorphismof HVJ virions.Virology 29:205-221. 10. Howe, C., C. Morgan,C.deVaux St. Cyr,K.C. Hsu,and

H.M. Rose. 1967.Morphogenesisoftype2parainfluenza virusexamined by light and electron microscopy. J. Virol. 1:215-237.

11. Hoyle, L. 1952. Structure of the influenza virus. Therelation between biologicalactivity and chemicalstructureofvirus fractions. J. Hyg. 50:229-245.

12. Kingsbury,D. W., and R. W.Darlington. 1968. Isolationand properties of Newcastle disease virus nucleocapsid. J. Virol. 2:248-255.

13. Kuhn, N. O., and C. G. Harford. 1963. Electron microscopic examination of cytoplasmic inclusion bodies in cells in-fected withparainfluenza virus, type2. Virology 21:527-530.

14. Reynolds,E.S.1963. Theuseof lead citrateathighpHasan

electron-opaquestain inelectronmicroscopy. J. Cell Biol. 17:208-212.

15. Rott, R., 1.M. Reda, and W. Schafer. 1962. Isolationand

characterization ofhemagglutinating non-infectious par-ticles produced during multiplication of NDV. Virology 16:207-209.

16.Weinberg, R.A., V.Loening, M.Willems, and S. Penman. 1967. Acrylamide gel electrophoresis of HeLa cellnucleolar RNA. Proc. Nat. Acad. Sci.U.S.A.58:1088-1095.

650

J. VIROL.