In vitro reassembly of vesicular stomatitis virus skeletons.

0022-538X/82/031055-08$02.00/0

In

Vitro Reassembly of Vesicular Stomatitis Virus Skeletons

WILLIAM W.NEWCOMB, GREGORY J. TOBIN,JOHN J. McGOWAN, ANDJAYC. BROWN* Department of Microbiology, UniversityofVirginiaSchool of Medicine, Charlottesville, Virginia 22908

Received14August 1981/Accepted 19 October 1981

Vesicular stomatitis virus (VSV) has been disrupted with nonionic detergent plus 0.5 M NaCl under conditions which result in solubilization of the viral glycoprotein (G), matrix protein (M), and lipids, leaving the nucleocapsid in a highly extendedstate.Dialysis of these suspensionstoremoveNaClwasfoundto

result in reassociation ofnucleocapsids with M protein. Reassociated structures were highly condensed and similar in appearance to "native" VSV skeletons produced by extraction of virions with detergent at low ionic strength. For instance, electron microscopic analysis revealed that, like "native" skeletons, "reassembled" skeletonswerecylindrical in shape, with diameters in therangeof

51.0to55.0nmandcross-striations spaced approximately 6.0nmapartalong the length of thestructure. Likenative skeletons, reassembled skeletonswerefound

by sodium dodecylsulfate-polyacrylamide gel electrophoresistocontain the viral N and M proteins, but they lacked the glycoprotein entirely. Both native and reassembledskeletonswerefoundtobe capable of in vitroRNA-dependent RNA synthesis (transcription). In vitro skeleton assembly required the presence of M protein and nucleocapsids. No skeleton-likestructureswereformed by dialysis of nucleocapsids in the absence of M protein or of M protein in the absence of nucleocapsids. These resultsprovidestrong supportfor the view that the VSV M protein plays afunctional role in condensing the viral nucleocapsid in vitro and raise the possibility that itmayplaya similar role in vivo.

Recent experiments in our laboratory have

emphasized the fact that NaCl can profoundly affect the disruption of vesicular stomatitis virus (VSV)bynonionic detergents(14).Extraction of native virions with60 mMoctylglucosideor1% Triton X-100 at low ionic strength results in solubilization ofthe viral glycoprotein (G) and

lipids, leaving insolublestructurescalled "skele-tons" (1, 12). Skeletons contain the viral RNA

complexed withthe N andMproteins in highly condensed andveryregularstructuresthat have

the same overall cylindrical shape and striated

appearance as the native virus. Similar struc-tures are producedby extraction with digitonin (18) or withTween 80 plus diethylether (2). In contrast, detergent extraction in thepresenceof 0.5 M

NaCl

results in solubilization of the G protein, theMprotein, and lipids,liberating the

viral nucleocapsid in an irregular and highly extended state. One molar KCl was found to have a similar effect on the extraction ofSendai virus with 2%TritonX-100 (15).

Thedramatic effectof NaClon the detergent-induced disruption of VSV has motivated us to ask whether its removal from detergent-high salt-solubilized virus might promote some

de-greeof reassociation of viralcomponents. Intact

vesicular stomatitisvirions were, therefore,

sol-ubilizedin nonionic detergent containing0.5 M

NaCl and then dialyzed against a

low-ionic-strength bufferto remove NaCl. The results of

this basic experiment, described below, show

that dialysis is accompanied by the reassembly of VSV nucleocapsids and M protein to form

condensed structures quite similar to "native" VSV skeletons.

MATERIALS ANDMETHODS

Cell and virus growth. All experiments were per-formed with the Mudd-Summers strain of VSV

(Indi-ana), which was grown on monolayer cultures of BHK-21 cells. Cellswere propagatedat 37°C in 150-cm2 plastic tissue culture flasks containing 40ml of Dulbecco modified minimal essential medium

supple-mented with10% calf serum,10%tryptosephosphate

broth, 100 U of penicillin per ml, and 0.1 mg of

streptomycinperml.Virus stocksweregrownat alow

multiplicity of infection as previously described (5)

and were shownbyrate-velocityultracentrifugationto

befree from defectiveinterfering particles.Viral pro-teinswereradioactively labeledbyincluding5p.Ciof

L-[3H]leucine (60Ci/mmol;Amersham Corp.)per ml inthe virusgrowth medium. VSV was purifiedfrom the infected cell medium by a two-step procedure

involving rate-zonal followed byequilibrium

ultracen-trifugation,essentiallyasdescribedbyHuntand Wag-ner(9). Thepurified viruspreparationswerejudgedto

befree from cellular contaminationby sodiumdodecyl

sulfate (SDS)-polyacrylamide gel analysis which

re-vealedonly thefive VSV proteins (N,NS, M, G,and

1055

on November 10, 2019 by guest

http://jvi.asm.org/

L); nocontaminating cellular proteins could be detect-ed.

Disruption of VSV. Freshly purified VSV to be extractedwith detergent was gently suspended in 0.01 MTris-hydrochloride buffer, pH 7.4, and centrifuged for 10 min at 1,000 x gto remove virus aggregates. Thesupernatant virus suspension was then extracted with either octylglucoside

(1-O-n-octyl-P-D-glucopy-ranoside;Calbiochem, La Jolla, Calif.) or Triton X-100 in thepresence of 0.5 MNaCl. Octylglucoside extrac-tion was carried out byadjusting the virussuspension to60 mM octylglucoside, 0.5 M NaCI, 5 mM dithio-threitol,10% glycerol, and 10 mM Tris-hydrochloride buffer, pH 7.8 (octylglucoside-0.5 M NaCl dissocia-tion medium) at0°C, and a concentration of 0.1 to 0.2 mg of viral protein per ml. Similar conditions were employed for Triton X-100extraction except that 1% Triton X-100 was substituted for 60 mM octylgluco-side. In bothcases virussuspensions were thoroughly butgently mixed with detergent plusNaCl and allowed tostand on ice(0°C) for 30 min before further opera-tions were performed. The initially turbid virus sus-pensions were clarified immediately upon addition of detergent plus NaCI, and the suspensions remained clearduring the 30-minperiod allowed fordisruption. Detergent extractionofVSV in the absence of NaCI wascarried out as describedpreviously (14).

Reassembly conditions. VSV components were al-lowed to reassociate during dialysis against 0.01 M Tris-hydrochloride (pH7.4)-S5mMdithiothreitol (TB). Four-milliliter samples of VSVdisrupted in detergent plus 0.5 M NaCI as described above were dialyzed against TB for 16 h at4°C and foranadditional 12 hat

25°C.Ifdialyzedsolutionswerenotanalyzed immedi-ately, theywerestoredonice foramaximumof 4 h. Ultracentrifugation. Insoluble materialwaspelleted

fromsuspensions of dissociated VSVorfrom reassem-bly mixtures byultracentrifugation in 5-ml Beckman SW50.1 nitrocellulose tubes. Four-millilitersamples of dissociated or reassembled viral components were layered on topof a 0.5-ml "cushion" ofglycerol and werecentrifuged at 38,000 rpm (130,000xg) for 2 h at

4°C.The fluid above andincludingtheinterfaceofthe glycerol pad wasaspirated while the remainder of the pad, containing thepelleted material,wasmixed with 0.5 mlofTBanddialyzedovernightat4°Ctoremove glycerol. Dialyzed samples were used directly for electron microscopic and transcription analyses or werelyophilizedprior toSDS-polyacrylamide gel elec-trophoresis.

Both rate-velocity and sedimentation equilibrium sucrose gradient methods were employed to purify

subviralstructures fromdissociated and reassociated VSVcomponents. Rate-velocity analyses began with 0.5 ml ofsample whichwaslayeredonto thetopof10 to70%(wt/vol)linearsucrosegradientspreparedin 5-ml Beckman SW50.1 nitrocellulose tubes. Stock su-crosesolutionswereprepared bydissolvingultrapure sucrose(Schwarz/Mann)inTB.Gradientswere centri-fuged at38,000 rpm for 90min at4°C inanSW50.1 rotor. After centrifugation, the gradients were frac-tionated from the bottomby use ofaperistalticpump andanLKB-7000 fraction collector.Theradioactivity

present in each fractionwasdeterminedby dissolving

a

20-piu

sample in 0.5 ml of NCS tissue solubilizer

(AmershamCorp.)containing10ml oftoluene-based scintillation cocktail(Research ProductsInternational,

Grove Village, Ill.) and counting in a Packard model 3320 liquid scintillation spectrometer. Gradient frac-tions to beanalyzed further were dialyzed against TB overnight at 4°C to remove sucrose and then were stained directly for electron microscopy or lyophilized for SDS-polyacrylamide gel analysis. Equilibrium ul-tracentrifugation was carried out by the same method described for rate-velocity centrifugation except that centrifugation was for 16 h rather than 90 min. The density of each gradient fraction was determined by weighing a100-,ul sample on a Mettler H18 analytical balance.

In vitro RNA-dependent RNA synthesis and poly-acrylamide gelanalysis of RNA. Native and reassem-bled skeletons were tested for the presence of endoge-nousRNA-dependent RNA polymerase activity by the assay ordinarily employed for intact VSV (4). Assay mixtures contained 0.14 M NaCl,0.2% TritonX-100,

7.5 mM MgCl2, 0.01MTris-hydrochloride, pH 7.9, 1 mMdithiothreitol, 0.3 mM ATP, GTP, andCTP, and

0.03mM

[a-32P]UTP

(0.45Ci/mmol) plus 25 to 35 ,g of

viral protein in a total volume of 30,lI. Incubations for up to 3 h werecarried out at 31°C, after which

5-RI

samplesof the reaction mixture were precipitated in 1 ml of5% trichloroacetic acid. Precipitated RNA was collected by filtration on 24-mm filters (Millipore Corp.), dried, and counted at an efficiency of 65% in toluene-based liquid scintillation cocktail. Results were expressed as nanomoles of[32P]UMP incorporat-ed per minute per milligram of protein.

Other methods. Electron microscopicanalysis was

carriedoutwith subviral structures negatively stained with2% phosphotungstic acid (pH 7.0) as described previously (14). Particle measurements were made on

positive enlargements of electron microscope nega-tives.SDS-polyacrylamide gel electrophoresis and gel staining with Coomassie blue were performed on 2-mm-thick slab gels as described by Nagpal and Brown (13).Seven-millimeter lanes were loaded with samples containing 50 to 100 ,ug of viral protein. All gels containedalane, labeled "std," for electrophoresis of solubilized VSV, andpositions of the five VSV pro-teins (L, G, N, NS, and M) areindicated. All protein concentrations were determined by the Lowry method (11).

RESULTS

Disruption ofVSV in nonionic detergentplus NaCl. Vesicular stomatitis virions were disrupt-ed innonionicdetergent

plus

0.5 MNaCl under

conditions shown previously to solubilize the viral envelope components and release the nu-cleocapsid in an uncondensed form (14). The extent of detergent-NaCl-induced virus

disas-semblywasmonitoredby electron

microscopy,

by sedimentation velocity ultracentrifugation,

andby

SDS-polyacrylamide gel electrophoresis.

Figure 1A shows a negatively stained

prepara-tion of insoluble structures

pelleted

from VSV

disrupted with 60 mM

octylglucoside-0.5

M

NaCI bycentrifugationfor 2 hat130,000x gas described in Materials and Methods. Pelleted material was found to contain only extended

viral nucleocapsids (6); no intact VSV or

con-densed VSV skeletons could be detected. It

J. VIROL.

on November 10, 2019 by guest

http://jvi.asm.org/

REASSEMBLY OF VSV SKELETONS 1057

C

FIG. 1. Electron micrographs ofdetergent-extracted VSV and reassembled VSV skeletons. (A) Nucleocap-sids isolatedbycentrifugationof solutionsproducedbyextractingVSV with 60 mMoctylglucoside plus0.5 M NaCl. (B) Skeletons isolated by centrifugation of suspensions produced by extracting VSV with 60 mM octylglucoside. (C) Reassembled skeletons (arrows) identifiedby direct observation ofthesuspension resulting fromdialysis of VSV solubilized with60 mMoctylglucoside-0.5MNaCl. (D) Reassembled skeletons isolated by

centrifugationof thesuspension resulting fromdialysis of VSV solubilized with60 mMoctylglucoside-0.5 M

NaCl.(A) x105,000;(B,C, and D) x125,000. Bar= 0.2 ,.m.

must be, therefore, that all or nearly all intact virus particles were disassembled.

Figure 2A shows the results obtained when

VSV,grownin the presenceof[3H]leucine,was disrupted in 60 mM octylglucoside plus 0.5 M

NaCl and analyzed by rate-velocity

sedimenta-tionon10 to70%sucrosegradients. Two bands

of3H-labeled viralproteins were observed, one

(bandII) at the top of the gradient where soluble

proteins should occur and the other (band I) VOL. 41,1982

on November 10, 2019 by guest

http://jvi.asm.org/

[image:3.491.54.446.69.538.2]

04

N

0 X 10

5

a.

B

I 10_

5

-1

Pellet

Bottom Top

FRACTION NUMBER

FIG. 2. Rate-velocity ultracentrifugation of VSV solubilized with 60 mM octylglucoside-0.5 M NaCI

before (A) and after (B) dialysis against TB. VSV employed in these experiments was radioactively la-beledby growth in the presence of[3H]leucine.

fartherdown the gradient. Electronmicroscopic

analysis

of bandIdemonstrated that it contained

extended viral nucleocapsids

indistinguishable

from those shown in Fig. 1A.

SDS-polyacryl-amide gel analysis, as shownin Fig. 3, demon-strated that band I contained predominantly N

protein (lane 2) whereas bandIIcontained both G and M proteins (lane 1). Virtually identical results of bothelectronmicroscopicand sucrose

gradientanalyses wereobtainedwhen1% Triton

X-100was substituted for 60 mM octylglucoside

in the virus dissociation medium. We conclude

that disruption of VSV in nonionic detergent plus0.5 MNaCl results in solubilization of the G and Mproteins and release ofthe viral nucleo-capsidin an extended state.

Reassociation ofVSV components: electron mi-croscopic analysis.Reassociation of viral

compo-nents wasassayed after vesicular stomatitis viri-ons were disrupted in nonionic detergent containing 0.5 MNaCl and then dialyzedagainst

TB. Dialysis was expected to remove detergent

andNaCl in the case of virus disrupted in octyl-glucoside-0.5MNaCl and NaCl onlyin the case of Triton X-100-0.5 M NaCl disruption (8, 17). Dialysis was found to promote reappearance of turbidity in initially clear solutions of viral com-ponents. Turbidity increased gradually during the first 12 to 18 h of dialysis and did not change thereafter. Direct electron microscopicanalysis

ofdialyzed suspensions revealed the presenceof condensed structures, as shown in Fig. 1C.

Extended nucleocapsids were very rare or ab-sent altogether. Condensed structures all

ap-peared to be made up of a fine filament or thread. In some cases the thread or threads

appeared tobe clumped in an irregular way to form roughly spherical particles ("tangled

clumps"),asshowninthe upper left of Fig. 1C.

Quite often, however, condensed particleswere

cylindrical in shape with cross-striations over the central portion of the overall structure and

irregularly packagedthreads at the ends (arrows in Fig. 1C). The cylinder diameter and the

Std1

2 3

4

L

G

urn

NS'

N

*

M9

-FIG. 3. SDS-polyacrylamide gelanalysisof subvi-ralstructuresseparated byrate-velocity

ultracentrifu-gationof VSV solubilized with60 mM

octylglucoside-0.5 MNaCl before(lanes 1 and2) and after(lanes3 and4)dialysis againstTB. Lane 1: bandII fromFig.

2A. Lane 2:bandIfromFig.2A. Lane 3:bandIIfrom

Fig. 2B. Lane 4: bandI fromFig. 2B. Thegel was stained with Coomassie blue, and the positions of standardVSVproteinareindicated.

on November 10, 2019 by guest

http://jvi.asm.org/

[image:4.491.61.234.68.422.2] [image:4.491.293.412.355.571.2]

spacing between cross-striations were found to

be similartothose observed for "native" VSV skeletons,asshownin Table1. Thelength ofthe

cross-striated region was variable but never greater thanthelength ofnative skeletons. The morphological similarity ofnative and reassem-bled skeletons was emphasized by electron mi-croscopic analysis of insoluble material pelleted from reassembly mixturesbyultracentrifugation

at 130,000 x gfor 2 h. Reassembled skeletons harvested in this way were indistinguishable

from native skeletons, as shown inFig. 1Dand B.

Control experiments showed that VSV

nu-cleocapsids and soluble components must be present together for striated skeletons to form

during dialysis.Theseexperiments involved dis-assembly of native virions in60 mM octylgluco-side plus 0.5 M NaCl as described above and

thencentrifugationto removethenucleocapsids from solution. The

supernatant

containing solu-bilized viral G protein, M protein, and lipids (soluble components) was then dialyzed against

TB under the same conditions employed for reassembly of viral skeletons. Similarly, the isolated nucleocapsids were resuspended at a

concentration of 0.1 mg ofprotein per ml in octylglucoside-0.5MNaCldissociation medium and dialyzed against TB. Electron microscopic analysis ofthe two dialyzed fractions revealed

thatneither contained skeletonsorskeleton-like

structures. Isolated nucleocapsids gave images similartoFig. 1Abothbefore and afterdialysis. Dialysis of soluble components resulted in for-mation of a precipitate which

pelleted

after

centrifugation

at600 x gfor 5 min.

SDS-poly-acrylamide gel analysis showed that the

precipi-tatecontainedMprotein only (datanotshown); theglycoprotein remained in the supernatant.

Protein composition of reassembled skeletons.

The protein composition of reassembled

skele-tons wasdeterminedby SDS-polyacrylamide gel analysis ofstructures purified from reassembly mixtures by sucrose density gradient

ultracen-trifugation.

Preparation of reassembled

skele-tons began with VSVgrown in thepresence of

[3H]leucine.

Purified virions were disrupted in nonionic detergent plus 0.5 M NaCl, and the resulting solutions weredialyzed toreassemble skeletons as described above. Reassembled

skeletons were purified from the dialyzed

sus-pensions by rate-velocity ultracentrifugation in one set of experiments and by sedimentation equilibrium in another. Figure 2B shows the

results of rate-velocity purification performed with skeletons reassembled after disruption of VSV with 60 mM octylglucoside plus 0.5 M

NaCl.

3H-labeled

viral components werefound

tosediment intwobands,one(band

II)

near the top of the gradient and the other (band

I)

cen-TABLE 1. Dimensions of native and reassembled VSVskeletonsa

Spacingof Skeletons Length (nm) Diam(nm) striations

(nm)

Native 204.4 ± 6.00 51.7 ± 3.00 5.7± 0.75

Reas-sembled

Variable'

53.7 ± 2.70 6.2 ± 0.32 a Results are reported as the mean value + 1 stan-darddeviation for at least 20 measurementsof

repre-sentativeparticles.Datafornativeskeletonsarefrom Newcomb and Brown(14).

bThe length of the cylindrical, striated portion of

reassembled skeletons was variable but nevergreater than 200.0 nm.

tered atfraction12.Electronmicroscopic

analy-sis revealed the presence of reassembled skele-tons inband I, but not inband II (micrographs

notshown). SDS-polyacrylamide gel analysis of band I from Fig. 2B (reassembled skeletons) demonstratedthat it contained the viral M and N

proteins, as shown in Fig. 3, lane 4. No G

protein was detected in reassembled skeleton

preparations; G proteinwas foundinband II at

the top of the gradient (see Fig. 3, lane 3). Results essentially identical to those shown in

Fig. 2 and 3 were obtained when 1% Triton

X-100wassubstituted for 60mMoctylglucoside in

thevirus dissociation medium.

When dialyzed reassociation mixtures were

centrifuged to equilibrium, two bands of

3H-labeled viral components were observed, one (band I) with abuoyantdensity ofapproximately

1.26g/cm3and theother(band II)nearthetop of the gradient. Electron microscopic analysis demonstrated thepresenceof reassembled

skel-etonsin bandI;none werefound in band II.

SDS-polyacrylamide gel analysis of band I material showed that M and N were the predominant

protein components present (seeFig. 4, lanes 1 and 3). As inthe case ofreassembledskeletons

purified by rate-velocity centrifugation, little or no G protein could be detected in reassembled skeletons purified by equilibrium sedimentation. Instead, G protein was found in band II at the

topofequilibrium gradients,asshown inFig. 4,

lanes 2 and 4. The identity of the disrupting detergent did not substantially affect eitherthe buoyant density or the protein composition (comparelanes 1and3inFig.4)ofreassembled

skeletons. Both parameters were virtually the samein skeletonsreassembledfrom virions dis-rupted with octylglucoside-NaCl and with Tri-tonX-100-NaCl.

RNA-dependent RNA polymerase activity. In vitro assays showed that both native and

reas-sembled skeletons containedendogenous

RNA-dependentRNApolymeraseactivity. Their spe-cific polymerase activities were found to be 41,

on November 10, 2019 by guest

http://jvi.asm.org/

[image:5.491.251.445.93.162.2]

Std 1 2

L

G

NS

N

_

4

M

d-FIG. 4. SDS-polyacrylami sembled skeletons and solubl by sedimentation equilibrit Lane 1: skeletons (band I) i

solubilized with60 mM octy Lane2:solublecomponents

reassembly of skeletonsfrom mM octylglucoside-0.5 M N (band I) reassembledfrom V Triton X-100-0.5 M NaCl. I nents (bandII) remaining aft tonsfromVSVsolubilized wi MNaCl. Gelswere stainedw

the positions of standard VSN

comparable to the activit nucleocapsids,asshown ir

tially lower than the spec

virions assayed under t] Polyacrylamide gel analys

patternofsimilarity in the sized by reassembled sk

tons, and native VSV. I RNA (3) and many

hip

RNA species were syi

shown).

DISCUSS The results reported I

formation ofVSV skelet( reassembly took place dui of solutionsproducedbyd

ularstomatitis virions inn

taining 0.5 M NaCl. The sembled under these con

be virtually indistinguisi VSV skeletons isolated b ruswithnonionicdeterger (14). For example, both n, skeletonswere foundtob

ders, with diameters of5

tions spaced approximatt native andreassembled si viral N protein, M proteil

3 4 Std protein entirely.

tons, reassembled skeletons were found to be

L

capable

of in vitro

RNA-dependent

RNA

syn-thesis. There can be little question, therefore,

thatreassembled skeletons arefaithful replicas

oftheir native counterparts.

It is most likely that removal of salt, not

removal of detergent, was responsible for

pro-G moting skeleton formation in vitro. Skeletons

oz1 i NS formedabundantly during dialysisof virus

solu-N

bilized with

octylglucoside-NaCl

when both

de-tergent and NaCl were removed. They also formed normally from virus solubilized with

so_ Triton X-100-NaCl when only NaCl was

re-_ -111 M moved. The presence of detergent, therefore,

did not appear to affect skeleton formation in

vitro, but thepresenceof NaCl clearly did. Salt de gel analysis of reas- wasfoundtohaveasimilareffectontheability

ecomponents separated of Semliki Forest virus glycoproteins to

asso-reassembledntfrfoumga iSnV

ciatewithSemliki Forest virus nucleocapsids in eglucoside-0.5 M NaCl. vitro. Glycoprotein-nucleocapsid interaction

(band II) remainingafter was observed in 24 mM octylglucoside-10 mM

VSV solubilized with60 sodiumphosphate buffer, pH 6.8,but not in the laCl. Lane 3: skeletons same solutioncontaining 0.1 M NaCl atpH 8.0

'SV solubilized with1% (7).

Lane 4: soluble compo- Althoughmatureskeletonswerethe

predomi-terreassembly of skele- nant structuresformedduringdialysis of

solubi-ith 1%Triton X-100-0.5 lized VSV, other related structures were also

ithCoomassieblue,and observedatlower frequency. These included the

Vproteinsareindicated. "tangled clumps" described above andtangled

clumpscontaining central cross-striations. Tan-ty observed for VSV gledclumps appearedtoconsist ofafinethread

iTable2, but substan- wound in an irregular way to form a roughly :ific activity of intact spherical particle. The thread diameter was he same conditions. smaller than the diameter of VSV nucleocap-is revealedan overall sids, and no nucleocapsids were found to

pro-RNAspecies synthe- trude from tangled clumps. Cross-striated

tan-eletons, native

skele-n all cases "leader"

gher-molecular-weight

nthesized (data not

SION

here demonstrate the ons in vitro. Skeleton ring overnight dialysis

lissolvingnative vesic-onionicdetergent con-skeletons which reas-ditions were found to

hable from "native"

y extracting intact vi-at vi-atlowionicstrength ,ativeandreassembled

)ecross-striated

cylin-il to 55 nm and

stria-ely 6 nm apart. Both

[image:6.491.74.231.81.231.2]

keletons containedthe n, andRNA, but both

TABLE 2. RNA-dependent RNA synthesis by native skeletons, reassembled skeletons, and

nucleocapsidsa

Specific % of native Preparation polymerase

activityb VSV

NativeVSV 1.78 100

NativeVSV skeletons 0.45 25

Nucleocapsids 0.44 25

Reassembled skeletons 0.38 21

(Triton X-100)

Reassembled skeletons 0.89 50

(octylglucoside)

a Native skeletons and nucleocapsids were prepared by disruption of VSV in 60 mM octylglucoside in the absence of salt and in the presence of 0.5 M NaCl, respectively, asdescribed previously (14). Reassem-bled skeletons were prepared by dialysis of VSV solubilized with Triton X-100-NaCl or octylglucoside-NaClasdescribed in Materials and Methods.

bExpressedasnanomolesof[32P]UMP incorporat-edper minute permilligram of protein.

J. VIROL.

I

on November 10, 2019 by guest

http://jvi.asm.org/

[image:6.491.259.450.465.590.2]

REASSEMBLY OF VSV SKELETONS 1061

gled clumps were similar in appearance to

tangled clumps except that they contained a central cylindrical region with regularly spaced striationsandtangledclumplikematerialatboth ends. The centralcylindricalregionwasof

vari-ouslengths indifferentparticles, but its diameter and striation spacing were the same as those found inmatureskeletons. Particles of thistype can be seen in Fig. 1C (arrows). We presume

that theminor components (tangled clumps and cross-striated tangled clumps) found in reassem-bly mixtures are intermediates in the skeleton self-assemblyprocess.Their structural

relation-shiptomature skeletons suggeststhat skeleton

reassembly invitro took place inthe following way. Nucleocapsids and M protein aggregated

to form tangled clumps in which condensed nucleocapsids weresurroundedbyafinethread

composed of M protein. Cross-striations were

then formed by subsequent

internal

rearrange-ments in the tangledclump structure. This

pro-posed pathwayforskeleton reassembly is shown

schematically

in Fig. 5. Further experimental

studies will be requiredtosubstantiate its role in skeleton reassembly.

The results described in this paper indicate that Mprotein musthave avery strong affinity for VSV nucleocapsids. Thiscanbeappreciated from thefact that veryfew,ifany,free nucleo-capsids remained after dialysis of virions solubi-lized with detergent-0.5 M NaCl;

virtually

all

were complexed with M protein in condensed

structures(see Fig. 2B).

Similarly,

nearly allM

protein became associated with nucleocapsids. Very little remained in the soluble fraction, as

shown inFig. 3,lane3, orFig.4, lanes2and4.

This contrasts tothe behavior of the glycopro-teinwhichwaspresentduringdialysis of solubi-lized VSV, butwas notincorporated into

reas-sembled skeletons. The interaction of nucleocapsids with M protein must, therefore, bevery strongandatleast somewhatspecific.

The abundantformation of VSV skeletons in vitro raises the possibility that similarstructures may be formed in vivo. Calculations based on the data ofSimonsen et al. (16) show that the molar concentrations of nucleocapsids and M

protein in

VSV-infected

HeLacellsare

compa-rableto or greater than their concentrations in

thereassemblyexperiments described here.

De-pending on the values assumed for the cell

volumeand the time after infection, the

concen-tration of negative-strand nucleocapsids in

in-fected cells

falls

inthe range of2 x

10-9

to 4 x

10-9

MandthatofM

protein

isapproximately 2 x

10-6

to 5 x 10- M. The concentrations employed forinvitro skeleton

reassembly

inour

experiments were 1 x

10-9

to 2 x 10- M for

nucleocapsids and 1.2 x

10-6

to 2.4 x

10-6

M

forMprotein.Theconcentrationsof

[image:7.491.253.445.79.109.2]

nucleocap-a. b. C. d.

FIG. 5. Proposed pathwayfor invitroreassembly

ofVSV skeletons.(a) Tangled clumps, (bandc)

cross-striated tangled clumps, and (d) reassembled

skele-tons.

sids and Mprotein are, therefore, high enoughto support skeletonformation in infected cells.

Experimentalresults ofKnipeetal. (10)also

suggest thatMprotein may associatewith VSV

nucleocapsids invivo. These authors examined CHO cells infected with a strain of

VSV,

tsM601, which carries a temperature-sensitive

lesion in the N protein. An unusuallylow level ofVSV nucleocapsids accumulated in cells

in-fected with tsM601 at the nonpermissive

tem-perature(39°C);Nproteinwasrapidlydegraded,

andlittlefull-length (42S)viral RNA was

synthe-sized. Incontrast,G proteinwasmade innormal amountsand itmigrated normallytothe

plasma

membrane. M protein was also synthesized in normal amounts, but it accumulated in soluble form in the cytoplasm rather than becoming attachedtomembranesasitdoesincells infect-ed with wild-type VSV. This is the result that

would be expected ifMprotein could associate

with membranes only after it had bound to

nucleocapsids.

ACKNOWLEDGMENTS

We thankNancySalomonskyandMargiePhippsfor techni-calassistance.

This workwas supported byNational Institutes of Health contractN01-NS-8-2391 and by Public Health Service grant HD-06390from the National Institutes of Health.

LITERATURE CITED

1. Cartwright, B., C. J. Smale,and F.Brown.1970. Dissec-tion of vesicularstomatitis virus into the infective ribonu-cleoprotein andimmunizingcomponents.J. Gen. Virol. 7:19-32.

2. Cartwright, B., P. Talbot, and F. Brown. 1970. The proteins ofbiologically active sub-units of vesicular sto-matitis virus. J. Gen. Virol. 7:267-272.

3. Colonno,R.J.,andA. K.Banerjee. 1976.AuniqueRNA speciesinvolved ininitiationof vesicular stomatitis virus RNAtranscription in vitro. Cell 8:197-204.

4. Emerson, S. U., and R. R. Wagner. 1972.Dissociation and reconstitution of the transcriptase and template activities of vesicular stomatitis B andTvirions. J.Virol. 10:297-309.

5. Fong,B.S.,R.C.Hunt, and J. C. Brown. 1976. Asymmet-ricdistribution ofphosphatidylethanolamine in the mem-braneofvesicular stomatitis virus. J.Virol. 20:658-663. 6. Heggeness, M.H., A. Scheid, and P. W. Choppin. 1980.

Conformation of the helicalnucleocapsids of paramyxo-viruses and vesicular stomatitis virus: reversible coiling anduncoiling induced by changes insaltconcentration. Proc. Natl.Acad. Sci.U.S.A. 77:2631-2635.

7. Helenius, A., and J. Kartenbeck. 1980. The effects of octylglucoside on the Semliki Forest Virusmembrane. Eur.J. Biochem. 106:613-618.

VOL. 41,1982

AI&

--

c.l

1-

-qw,T

on November 10, 2019 by guest

http://jvi.asm.org/

8. Helenius, A., and K. Simons. 1975. Solubilization of membranes by detergents. Biochim. Biophys. Acta 415:29-79.

9. Hunt, D. M., and R. R. Wagner. 1975. Inhibition by aurintricarboxylic acid and polyethylene sulfonate of RNAtranscription of vesicular stomatitisvirus.J.Virol. 16:1146-1153.

10. Knipe, D. M., D. Baltimore, and H. F. Lodish. 1977. Maturation of viral proteins in cells infected with tempera-ture-sensitive mutants of vesicular stomatitis virus. J. Virol. 21:1149-1158.

11. Lowry, 0.H., N. J. Rosebrough, A. L. Farr, and R. J.

Randall.1951. Proteinmeasurementwith the Folin phenol

reagent.J.Biol. Chem.193:265-275.

12. Miller,D. K., B. I. Fever, R. Vanderoef, and J. Lenard. 1980.Reconstituted G protein-lipid vesiclesfrom vesicu-lar stomatitis virus and their inhibition of VSV infection. J. Cell Biol.84:421-429.

13. Nagpal, M. L., and J. C. Brown. 1980. Protein and glycoprotein componentsof phagosome membranes

de-rivedfrommouseLcells. Int. J. Biochem. 11:127-138.

14. Newcomb,W. W., and J. C. Brown. 1981. Role of the vesicular stomatitis virus matrix protein in maintaining the viral nucleocapsid inthe condensed form found in nativevirions. J. Virol. 39:295-299.

15. Shimizu, K., and N. Ishida.1975. Thesmallest protein of Sendai virus: its candidate function of binding nucleocap-sidtoenvelope. Virology 67:427-437.

16. Simonsen, C.C., S. Batt-Humphries, and D. F.Summers. 1979.RNAsynthesis of vesicular stomatitis virus-infected cells: in vivoregulationofreplication.J. Virol. 31:124-132.

17. Stubbs, G.W.,H.G. Smith, Jr.,andB.J.Litman.1976. Alkylglucosidesaseffective solubilizingagentsforbovine rhodopsin: a comparison with several commonly used detergents. Biochim. Biophys. Acta 426:46-56. 18. Wagner,R. R., andC.A.Schnaitman. 1970. Proteins of

vesicular stomatitis virus,p.655-671.In R. D.Barryand B. W.J. Mahy (ed.),ThebiologyoflargeRNA viruses. AcademicPress, Inc.,New York.

on November 10, 2019 by guest

http://jvi.asm.org/