Maturation of viral proteins in cells infected with temperature-sensitive mutants of vesicular stomatitis virus.

(1)

Maturation of Viral Proteins in Cells Infected with

Temperature-Sensitive Mutants of Vesicular Stomatitis

Virus

DAVID M. KNIPE,' DAVID BALTIMORE,* AND HARVEY F. LODISH

Department of Biology, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139

Received forpublication 10 September 1976

Maturation ofviral proteins in cells infected with mutants of vesicular stoma-titis virus was studied by surface iodination and cell fractionation. The move-ment ofG, M, and N proteins to the virion bud appeared to be interdependent. Mutations thought to be in G protein prevented its migration to the cell surface, allowed neither M nor N protein to become membrane bound, and blocked

formationof viral particles. Mutant G protein appeared not to leave the endo-plasmic reticulum at the nonpermissive temperature, but this defect was

par-tially reversible. In cells infected with mutants that caused N protein to be degraded rapidly or prevented its assembly into nucleocapsids, M protein did not bind tomembranesand G proteinmatured to the cell surface, but never entered structures with the density of virions. Mutations causing M protein to be degraded prevented virion formation, and G protein behaved as in cells infected by mutants in N protein. These results are consistent with a model of virion

formationinvolving coalescence of soluble nucleocapsid and soluble M protein with G proteinalready in the plasma membrane.

The pathways of maturation of the major structural proteinsof vesicular stomatitis virus (VSV) have been characterized (5). The

glyco-protein (G) isboth insertedintothe membrane ofthe endoplasmicreticulum andpartially gly-cosylatedsorapidlythatintermediatesare not

evident. Thefinal stagesof glycosylationoccur

afterthe Gprotein has migratedto

light-den-sity membranes; soon thereafter it appears on

the surfaceof the cell. Atearly times afterits

synthesis, the M protein is soluble; from this state itisprogressivelyincorporatedinto

mem-branous structures with the density of virions and thenquicklyappears inextracellular

viri-ons. Thenucleocapsid (N) protein is also

solu-bleafter its synthesisand is laterincorporated

intonucleocapsids that attach to the membrane prior tobudding into extracellular virions.

Temperature-sensitive mutants of VSV have

beenisolatedbyseverallaboratories (1, 2, 4, 10, 11), andmutants in some of the

complementa-tionclasses have been shown tobedefectivein certainmajor structural proteins (7-9). To char-acterize the effects ofthese mutations on mor-phogenesis of virions, we have examined the virus-specific structures that accumulate in cellsinfectedwith thesetemperature-sensitive mutants atthenonpermissive temperature.

1Present address: CommitteeonVirology,Universityof Chicago, Chicago,IL 60637.

MATERIALS AND METHODS

The origin and growth of the virus strains, cell fractionation, lactoperoxidase-catalyzed iodination ofthe cells, and othermethods used in thispaper have been described previously(5-7).

RESULTS

MigrationoftheGprotein tothe surface of

cells infectedwithtemperature-sensitive mu-tants.Itwaspreviouslyshown that the G pro-teinoftsM501(V) didnotappeartoundergothe final step(s) ofglycosylation because the

elec-trophoretic mobility of the proteindid not de-creaseduringachaseperiod (7).This

change

in

mobilityhasbeen showntoinvolve the addition of sialic acid residues to the molecule several

minutespriortoitsappearanceonthe cell sur-face (6). We therefore tested whether the pro-tein encoded bythe mutant virus migrated to

thesurfaceof infectedcells. Thiswastestedin twoways,i.e.,surfaceiodination with

lactoper-oxidase and protease treatmentof [35S]methio-nine-labeled cells.

(i) iodination.The surfaceproteins of mock-infectedChinese hamster ovary cells labeled by lactoperoxidase-catalyzed iodination showed a pattern similartothat observedpreviously,

ex-cept that in this case some labeled protein,

migrating with bovine serum albumin and slightly slower than the VSV G marker, was 1149

on November 10, 2019 by guest

http://jvi.asm.org/

(2)

1150 LODISH

observed(Fig. 1). Thiswaspresumablythe re-sultof the serumproteinstickingtothe cells. In the cells infected with wild-type virus, a new

iodinated protein comigrating with virion G wasobserved onthe surface of the cells under allinfectionconditions (seethelegendtoFig.1

for an explanation of infectionconditions),and

thegreatest amount waspresent oncells kept

continuously at390C.

Iodination of cells infected with tsM301(III)

ateither 39 or 31'Crevealedthe presence of G protein on the surfaceofthe cells, but with a

greater amount at 390C than at 31'C (Fig. 1). These results indicatedthatincellsinfectedat

the nonpermissive temperature with a virus having an apparent mutationintheMprotein, the G protein maturednormallytothesurface

of the cells. It isofinterest to note that the G protein on the surface of cells infected with

rno

tsM301(III) at either 39 or 31'C did not comi-grate with the wild-type G protein, but mi-grated more slowly. This was previously ob-served with [35S]methionine-labeled proteins and isprobably due to differencesbetween our wild-type VSV and the parent of mutant tsM301(III) (7).

Themutantts045, agroupVmutantwhich, like tsM501(V), encodes aGproteinwhich

re-mains underglycosylated at 390C (data not

shown),waspreviously showntobedefectivein maturation of the G protein fromdense

mem-branestolight membranes (8). Weobservedno

Gprotein on thesurface of cells infected at 390C with ts045(V), but cells infected at 31'C

showeda considerable amount (Fig. 1). More-over,when cellsinfectedat390Cwereshiftedto

310C for only 1 h, a significant amount of G proteinaccumulated on the surface of thecells,

wt

M3O0(m)

045.-3

30 9 'S3

,

,mretne

.f

M 5)

1

... ... !.)

:rt ..

rne,

t

)r

9m

Jne

FIG. 1. Surfaceiodination ofcellsinfected with temperature-sensitivemutants.Cells wereinfected with the indicated virus strain at a multiplicity of 10. At the termination of the infection, the cells were washed and iodinated as described in the text. The total cellular proteins were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Protocol for infections:39°C,entireinfectionat39C-cells wereharvested

at5 hpostinfection; 31°C,entireinfectionat31°C- cells were harvested at 5 h postinfection; 39°C-- 31°C,

infectionat39°Cfor the first4h ofinfection and then shifted to 31°C for 1 h of incubation; 39°C-) 31 °C +

emetine, infectionat39°C for the first4 hof infection and then shifted to 31°C for 1 h of incubation in the presenceof 20 pg ofemetine perml (exposure time, 48 h).

http://jvi.asm.org/

[image:2.501.71.452.275.578.2]

(3)

showing that the defect was partially

reversi-ble. The appearance of G on the surface also

occurredinthepresenceof 20

gg

of emetine per

ml, aconcentrationsufficient to abolish all vi-rus protein synthesis. Hence, some of the G protein previously synthesized at 390C, but blocked in maturation within the cell, could move to thesurface of the cell once the temper-ature was lowered.

In addition, no labeled G protein could be found on the surface of cells infected with tsM501(V) at 390C, whereas at 31'C there wasa

detectable amount of G protein on the surface (Fig. 1). A small amount of G protein could be detected on the surface of cells within 1 h after cells, growing at 390C, were shifted to 31'C. This appearancealso occurred in the absence of protein synthesis, indicating that the defect waspartiallyreversible. Theincreasedlabel in the bands of the sample with emetine was

prob-ably dueto adarkerbackgroundbecause itwas

not apparent in other experiments. Thus, we

conclude that these two group V mutants are

defectivein maturation of the Gproteintothe surface of infected cells at the nonpermissive

temperature.

(ii) Protease treatment. To corroborate the

findings with lactoperoxidase labeling, we treated intact, infected cells with chymotrypsin toassay for the presence of the Gprotein on the cell surface. Infected cells were labeled with

[3:S]methioninefor 15 min and then incubated for an additional 60 min after the addition of excessunlabeled methionine. The infected cells werethen treated with chymotrypsin and pre-pared for gel electrophoresis. After a 60-min chase period, the G protein labeled at 39 or 31'C in cells infected with wild-type VSV was largely sensitive to prQtease treatment and thus was on the surface of the cells (Fig. 2).

The G proteinlabeled at 390C in cellsinfected

with tsM501(V) was not susceptible to protease treatment and thus not on the surface of the cells. Asevidencedby its protease sensitivity, it waspresent on the surface of the cells at 31'C. Thisconfirmed the conclusion from the

iodina-tsM

501(V)

310

39

31 tsM

301(I)

39

31 tsM 601

()

39

31

- _

_-_ _ _ _

+ - + - + + - + - +

Chymotrypsin

treatment

FIG. 2. Protease treatment of intact cells infected with temperature-sensitive mutants of VSV. Cultures

wereinfected with theindicated virus strainatamultiplicityof10 and incubatedat31°C for5 h. At that time thecellswereresuspendedincomplete medium lackingmethionine,andone-halfofthemwereplacedat39°C and theremainderwereplacedat31'C.Aftera10-minwarming period,thecultureswerelabeledfor15min with[35S]methionine. The cultures were then incubated withexcess unlabeled methioninefor60min. The cellswerewashedand treated withchymotrypsin(1 mgofchymotrypsinpermlfor10minat37°C, except for

tsM301(III), whichwastreated with 10mgof chymotrypsinperml). Exposuretimes: wt, 24h;tsM501(V), 72 h; tsM301(III), 24h; tsM601(VI),24 h.

wt VSV

39

0

G2N-NI

-Ns-

~~ - - 3

-

http://jvi.asm.org/

[image:3.501.50.444.316.579.2]

(4)

tion results that the G protein encoded by tsM501(V) didnot mature properlyto the

sur-face of the cellsat390C.

Thistechnique was also applied to cells

in-fectedwith temperature-sensitivemutantsthat

weredefective in RNA synthesisatthe

nonper-missive temperature. RNA synthesis was

al-lowedto proceed at31'C, and the culture was

then shifted to 390C prior to labeling with

[35S]methionine. Thus,all of the labeled G

pro-tein would be synthesized and mature at the nonpermissive temperature. In contrast,

sur-face iodination would not distinguish newly synthesized G proteinfromGproteinputonthe

surface priortothe temperature shift. In cells infected with tsM601(VI) at 39°C, alarge

per-centage of the G protein wasremovedby

pro-teasetreatmentand thuswas onthesurfaceof

the cells. Because replication of viral RNA is low in these cellsat39°Cand alsothe Nprotein isdegradedrapidly(7),therecould beat mosta

small pool of nucleocapsids in these cells at 39°C. Therefore, it appears probable that

nu-cleocapsids have little role in maturation of the Gproteintothe surface of cells.

When cells infected with tsM301(III)at39or

31°Cwereexposedto1mgof chymotrypsinper

ml for 10 min, only 10 to 20% of the

[35S]-methionine-labeledG protein or iodinatedcell

surface G protein was removed (data not shown). Treatment of cells with 10mgof

chy-motrypsin perml for the sametime, however,

did remove [35S]methionine-labeled G protein

(Fig. 2) aswell asthe iodinated cell surface G

40

30

2 0

;.0

M protein N protein G protein

wild typeVSV 39,

30' label 30' chose

Ls

Hn

n

-o I234 56 S

Gradient U

ructions P

a

,ln ,f1

23456SV

U P R

U

S

23456 S V

U P R

U S

protein (data not shown). Removal of the G protein from cells infected by the Glasgow strains ofVSV also required 10 mgofenzyme

per ml, suggesting that the parent of

tsM301(III)is the Glasgow strain and

support-ing the previousconclusion that tsM301(III) is the same mutant as tsG33(III). Using the

higher enzyme concentration, it is apparent

thatamutation of theMproteingenedoes not affect movement of the G protein to the cell

surface.

Fractionation of cellsinfected with

temper-ature-sensitive mutants. To examine the

ef-fectsof various mutationsonthe maturationof

the nonmutant proteins, we utilized the

frac-tionation procedure described previously (5). Cellswereinfectedwithwild-type and

temper-ature-sensitivemutantstrainsofvirus, labeled with[35S]methionine for30minat5h postinfec-tion, and subjected to chase conditions for 30 min at the permissive or nonpermissive tem-perature.

The proteins from each subcellular fraction

wererecoveredandsubjectedtosodiumdodecyl sulfate-polyacrylamide gel electrophoresis. The amountof eachviral proteinineachsubcellular fraction was expressed as apercentage ofthe

total viral proteinsinthe culture (Fig.3, 4,and 5).

(i) Cells infected with wild-type VSV. The

distribution of viral proteins in cultures in-fected with wild-type virus at 39 or 31'C was very similar to that described previously (5)

(Fig. 3). TheMprotein wasfound in the

cyto-M protein N protein G protein

4')

tsM30l(z) 39°C

30_ 30' lobel

30'chose

20 2

'O,0 .. _LE,,

,

,..--23456 S V 23456 S V

Gradient U U

fractions P U P UR

S S

40 40

wildtypeVSV tsM301(z)

z ~~~~~~~310 311C

30-30'label 30 30'lobel

30'chose 30'Chose

20--

20-0 _rF~,

Ln

z~oH

Xn

123456 S V 123456 S V 123456 S V 123456 S V 123456 S V 23456 S V

FIG. 3. Fractionation of cells infected with temperature-sensitive mutants of VSV: wild-type and tsM301 (III)proteindistributions. Cultureswereinfectedwith the indicated virus strainatamultiplicity of10

at39or31'C. At5 hpostinfection the cellswereresuspendedin complete medium lacking methionine and

labeledfor 30 min. Excessunlabeled methioninewasadded and incubationwascontinuedfor 30min. The

cellswerethen fractionatedby the procedure described byKnipeetal. (5), with theuseof0.1M NaCl inthe initial centrifugationtoeliminate Maggregation. Theproteinswererecovered fromeach subcellular fraction

andsubjectedtosodium dodecylsulfate-polyacrylamide gel electrophoresis. Theamountof each protein in eachfractionwasdeterminedand expressedas apercentageof the totalamountof viral proteins in the culture.

_, .. _: "

I

20

v

R U S

123456 S V

u

P R u

s

http://jvi.asm.org/

[image:4.501.117.404.424.580.2]

(5)

M protein N protein G protein Mprotein N protein G protein

FIG. 4. Fractionation of cells infected with temperature-sensitive mutants of VSV: tsM501(V) and

tsM601(VI). Theseexperimentswereperformedasdescribedinthe legendtoFig.2,exceptthat cells infected

with tsM501(V) andtsM601(VI)wereused. The culture infected withtsM601(VI), whichwaslabeledat390C,

wasincubatedat31°Cfor thefirst5h of infection and shiftedto390C 10minpriortolabeling.

30 201

Z3 4 55 S V Z 5 4 5 b 5 V Z345 6 S V 2 345 6 5 V 2345 6 S V 2 3 4 5 6 S V

FIG. 5. Fractionation of cells infected with temperature-sensitive mutants of VSV: tsG41(IV) and tsM502(V). These experimentswereperformedasdescribedinthe legendtoFig.2,exceptthat cells infected withtsG41(IV)andtsM502(V)wereused. The cultures labeledat390Cwereincubatedat31'C for the first3h of infection, and one-halfwere shifted to 390C for 2 h of incubation prior to labeling atthe appropriate temperatures.

plasmicsupernatantand inasymmetrical

dis-tribution about fraction 4of the isopycnic

gra-dient. The membranousstructures infractions 3 and 4that contain Mproteinarebelieved to

benewlybudded virus attachedtothe cellsand buddingintermediates. Theonlydifference be-tweencultures infectedat39 and31'Cwasthe

observation that the percentage of soluble M protein wastwice as great at31'C asat390C, possibly the result of the increased degradation rateof theMproteinat390C (7). The Nprotein

wasfoundlargelyinnucleocapsidsinfraction 1

atboth temperatures.Nproteinwasalso found inthe cytoplasmicsupernatantandinasecond

peak of N protein-containing structures

cen-tered around fraction 4 in the gradient

mem-brane fractions. After thislongchaseperiod the

G protein is found mainly in the subcellular

fractions enriched inplasma membranes, i.e.,

fractions 5 and 6. There was alsosome G

pro-tein in fraction 4, presumably in virions at-tachedtocells.

(ii) Mutations in the M protein. The distri-bution of viral proteins in cells infected with tsM301(III) at390C showed that the Mprotein

wasfoundinsimilar fractionson theisopycnic

gradient as in wild-type virus-infected cells,

withapeakcentered around fraction4 (Fig. 3).

However,therewasverylittle Mproteininthe

cytoplasmicsupernatant, whichmayhave been

aresult of therapid degradationrateof the M protein at 390C (7). The N proteinwas found

almostentirelyin freenucleocapsids,and much less Nprotein floatedupintogradientfraction

tsG 4(IY)

30 31' - 39-C

I prior to

30'label

27 30' chose

0

~H

171 n--,

n

-

fh

2345 6 5 V 23 456 5 V 2 3456 S VI

toM502(Y)

so0_ 31,-_39-C _

30' label

20 ~~jchose30' j

° 2 3 4 5 6 5 V 23 4 5 6 5 V 2 3 4 56 S V

tsG41(I()2)

31' J 30' label at

30' chase

nHn

h

n Jrni

ts M502 (7)

31-30'label

30'chase

rflte ..n11 H 11 rI..

401~ 12 401

--0._ A I**1*-*III**

I

2

7

I . . I.' .. - - . I. .. . 11

G,.d,;,, u u . ; z G'ad'" u u 4 u 4 R

p p p ? -; act.... p p p

s s s w

http://jvi.asm.org/

[image:5.501.100.392.69.234.2] [image:5.501.99.389.289.432.2]

(6)

1154 BALTIMORE, LODISH

4ascomparedwith cellsinfected with wild-type virus at 390C or tsM301(III) at 31'C. The G protein was found largelyin fractions 5 and 6, apparently in the plasma membrane, because ofthe previous experiments (5) demonstrating the sensitivity of the G protein to protease treatment of intact cells. In addition, there werevirtually no viral proteinsfoundin

extra-cellular virus particlesproduced by these cells. Thus, thismutation prevented the formation of physical virus particles as well as infectious virus. This was consistent with the observa-tions of Unger and Reichmann (13), which showed that cells infected with tsG31, another group III mutant, incorporated labeled RNA intonucleocapsids, but not into virus particles. In cells infected withtsM301(III) at 31'C the distributions of the viral proteins were very similartothose incellsinfected withwild-type

virus. Much more Mprotein wasfoundin the cytoplasmic supernatant than at 390C, and much more Nproteinfloated up with the mem-branes inthe isopycnic gradient.

(iii) Mutationsin theglycoprotein. In cells infected with tsM501(V) at 390C, we observed that the largest amount of G protein was in

fractions 1 and 2, the fractions shown to be richestinendoplasmic reticulum (Fig. 4).Thus, the G protein may be defective in movement

from the endoplasmic reticulum. Consistent with this observation are the findings that at

390C only the G1 form of the glycoproteinwas

found and that noneof the G proteinwasfound

onthe cell surface (Fig. 1and 2). The N protein

was found mainly in fractions 1 and 2as free nucleocapsids. TheMprotein wasfound largely

in the cytoplasmic supernatant, with smaller amounts scatteredacrossthe densitygradient.

The amount of M protein in the supernatant has varied from 55 to 67% in several experi-ments, whereas only 20 to 30% of the wild-type protein was present in the cell supernatant. Therefore, the presence of G proteinonthe cell surface appears to have a role in the attach-ment oftheMprotein to membranes. At 31'C theGproteinmigratednormally to the plasma membrane fractions,andtheMand N proteins were found intheirnormal distribution about fraction 4. Total labeled protein accumulation

in extracellular virus was at least fivefold higher at 31'C, indicating a defect in particle maturation at thenonpermissive temperature. Although the data are not shown, similar

pro-tein distributions have been observed for tsO45(V), amutant having asimilar defect in maturationofthe Gprotein.

(iv) Mutations affecting the N protein.

Mu-tant tsM601(VI) was shown to encode an N

protein thatdegrades rapidly at 39°C, and the

mutantsynthesizes little 40S viral RNA, even at 31°C (7). Since this mutant is defective in RNA synthesis, we allowed RNA synthesis at 31°C for5handthen shifted thecultureto 39°C

prior to labeling with [35S]methionine. The amountofnucleocapsidsinthe cellswas there-fore determinedbythesize of thenucleocapsid poolat31°C.Asexpected,theamountoflabeled

Nprotein inthe cells labeled at39°C wasvery low due to its rapid degradation (Fig. 4). The G protein inthese cellswas foundlargely in the gradient fractions that contain plasma mem-brane (fractions5and6). However, the M pro-tein was largely in the cell cytoplasmic super-natant, indicating that it cannot stably attach

tomembranes.The percentage offree Mranged

from 67 to 75%, and the remainder ofthe M

protein was distributed across the gradient, withslightly larger amounts closer to the top of thegradient (fractions 5 and 6). At 31°C the N protein was partly in nucleocapsids, but, as

compared with wild-type VSV, an unusually large proportionwasin thecytoplasmic super-natant. This was not unexpected since the

amount of viral RNA replication was low at 31°C, but the N protein was stable. Thus, lessN

protein was bound to RNA inthese cells. The amount of soluble M washigh,presumably due

tothelowrateofvirus assembly because of the lackofnucleocapsids and the lower rate ofM

proteindegradation at31°C.

Althoughthe amount of viral proteins

assem-bled into virus from cells infected with tsM601(VI) and labeledat 39or 31°Cappears to be the sameasthatshowninFig. 4,inabsolute

amountsthe proteinsat31°Cwere two- to three-foldhigher dueto thedegradation of virtually all ofthe N proteinand significant amounts of

Mprotein at39°C.Thismadethepercentagesof

viral proteinsin extracellular particles at 39°C artificially high. We presume that cells infected with tsM601(VI) have a low level of intracellu-lar nucleocapsids at 31°C, which leads to the lowlevel of virusformation observed.

Incells infected with tsG41(IV) and labeled

at 39°C, a reduced amount of N protein was

presentinthe cells duetoitsrapiddegradation (Fig. 5). TheundegradedN proteinwas found

inlow amounts infraction 1,possiblyin nucleo-capsids, but much was in the cytoplasmic

su-pernatant. The Gprotein was largely in frac-tions 5 and 6 and thus presumably in the plasmamembranes. In contrast tocells infected

with tsM601(VI), a high percentage of the M protein and some G protein were in membra-nous structuresdistributedaboutfraction 4on

the isopycnic gradient, with normal levels in the cytoplasmic supernatant. At 31°C a

three-tofourfold-higher amountof viralproteinswas

http://jvi.asm.org/

(7)

incorporated into extracellular virus, and the distributions of the viral proteins were very

similartothose of cells infected withour

wild-type virus.

We have included in this sectionexperiments describing the fractionation of cells infected withtsM502(V)because it is obvious from Fig.5

that tsM502(V) has a defect affecting the

as-sembly of N protein into nucleocapsids. In cells labeled at390C nearly all of the N proteinwas

presentinthe cytoplasmicsupernatantaswell

asmostof the viral M protein. Thus it appeared that the M protein could not stably bind to membranes in these cells. The G protein mi-grated normally to fractions 5 and 6, but seemedto represent alower percentageof the

total labeled viral proteins than normalatboth this temperature and 31'C. At 31'C a more

normal distribution of viral proteins was

ob-served, except that the amount of soluble N protein was still higher than wild type. This could mean that the defect is still partially

manifest even at the permissive temperature. Inspite of the fact thattsM502(V)wasassigned

to group V on genetic grounds, we have been

unable tofind anydefect in the maturation of

the G protein. The onlyapparentdefect is lower

levels of synthesis of G protein insome

experi-ments.

Temperature-shift experiments ofcultures infected with temperature-sensitive mutants. To explore whether the blockstovirion matura-tion defined above were reversible or not, we

incubated infected cells for 3 hat31°C, followed by2hat39°C, and the cellswerethenlabeled

for 30 min at 39°C with [35S]methionine. The cultureswerefurtherincubated withexcess

un-labeled methionine for 30 min. At that time one-halfof each infected culture wasremoved

and placedat0to4°C, and the remainderwas

shifted to 31°C for 1 h. At that time the cells

wereharvested from the cultures and their

cy-toplasm was fractionated into a supernatant

andapellet of particulate material,preparedas

in the previously described fractionation scheme. Theonly infected culture that incorpo-ratedasignificantamountof labeled viral

pro-teins into extracellular viral particles at39°C

was the culture infected with wild-type virus

(Table 1). After the shift intemperature to31°C the culture infected with wild-type virus

re-leased 2.5 times as much labeled protein in

extracellular virions. However, many of the cultures infected with temperature-sensitive

TABLE 1. Effectoftemperatureshiftonthe subcellular locationofproteins encodedbytemperature-sensitive

mutantsa

%of total:

M protein Nprotein Gprotein

Wildtype or mutant

Cellpel-

Cels-Cell

pel-

Cels-Cell

pel- Cell

su-let perna- Virus let perna- Virus let perna- Virus

tant, tant tant,

wt, 390C 38 18 44 69 12.5 18 90 0 10

wt, 390C -*310C 15 7 78 42 12 46 45 45

tsM301(III), 390C 93 7 0 93 6 1 99.6 0 0.4

tsM301(III),390C- 95 0 5 76 13 10 88 0 12

310C

tsM501(V), 390C 29 71 0 81 19 0 100 0 0

tsM501(V), 390C - 54 42 4 90 6 4 100 0 0

310C

tsO45(V), 390C 49 51 0 19 28 3 100 0 0

tsO45(V), 390C - 64 11 24 75 16 9 89 0 11

310C

tsM601(VI), 390C 36 64 0 16 83 0 100 0 0

tsM601(VI), 390C- 36 41 22 15 85 0 97 0 3

310C

tsG41(IV), 39°C 69 26 6 49 51 0.7 94 3 3

tsG41(IV), 390C 51 12 37 83 15 2 90 0 10

310C

aCultureswereinfected with the indicated virusandincubatedat310Cfor3h. Atthattimethe cultures

wereshiftedto390Candfurtherincubated for2h. The cultureswerelabeledfor30minwith[35S]methionine and thenincubatedwithexcessunlabeledmethionine for 30 min. At that timeone-halfof the culturewas

removed andplacedat0to40C.Theremainderwastransferredto310C and incubationwascontinued for 60 min.The cellswerefractionatedintoacytoplasmicsupernatant anda100,000xgpelletof membranous and particulate material. Centrifugation wasperformedinthe presence of 0.1 M NaCl toeliminate Mprotein aggregation.

http://jvi.asm.org/

(8)

mutant virus showed a much more dramatic increase intheamountof viral proteinsin

ex-tracellular virus,presumablydueto areversal of the temperature-sensitivedefect.

In cells infected with tsM301(III) the total

amount oflabeled viralproteins in virions in-creased approximately 20-fold after shifting to 31'C. Theamountof N andGproteinincreased

dramatically in extracellular virions after the shift-down intemperature, but very little of the

M protein that had not been degraded was

chased into virus, a result first observed by Lafay (8)fortsO89(III). Thus, Mprotein made

at 390C in cultures infected with tsM301(III)

was eitherin an aberrant structure or hadan

irreversibledefect. In a similarexperiment,the

extracellular virus titer increasedfrom5 x 104 PFU/mlinthe390C culture to8 x 106 PFU/ml

inthe infectionshifted to310Cfor 1h.Thus, it

appeared that infectious particles were

pro-duced after the shift intemperature, but little

M protein labeled at 390C was assembled into virus particles. This is evidence thatthe tem-perature-sensitive mutation of thetsM301(Ill) virus is adefect in theMprotein, asfirst noted by Lafay (8) forts089(III).

In cells infected with tsM501(V), the large

amountofsolubleMproteindecreasedafter the shift to 31'C, and a higher percentage of this protein wasboundtomembranes (Table 1). The G protein alsounderwent adecrease in mobil-ityas a resultofthe temperature shift, proba-bly a reflection of further glycosylation (not

shown). However, onlyasmall amountof viral protein assembled into extracellular virions

after thetemperatureshift,andthus thedefect

wasnotveryreversible.

The mutant tsO45(V), however, showed a more reversible phenotype as observed by

La-fay(8).After thetemperatureshift,the amount ofMbound to membranes increased, and the

amount of viral protein in virions showed an

eight- to tenfold increase. However, all of the labeled viral proteins were incorporated into virus in approximately the normal ratios,

in-cludingthe Gprotein,whichisblockedin mat-uration within the infected cell at 390C. This reversiblephenotype of theGprotein was also observed in the iodination experiments de-scribed above.

Incellsinfected withtsM601(VI) the

temper-atureshiftallowed anincrease ofincorporation of labeled proteins into virions. The small

amount of N protein in these cells was not incorporated into virus, but the M and G

pro-teins were. After the shift in temperature to 31°C wealso observed a decrease in the amount ofsoluble Min these cells.

The defect in virion maturation in cells in-fected with tsG41(IV)wasalso reversible. After the temperatureshift,alargeincrease in viral

protein assembly into extracellular virions oc-curred. ThelargeamountofMproteinbound to membranes in these cells decreased after the temperatureshift,andatleastpart of this ma-terialwasincorporatedintoextracellular virus.

Verylittle of theundegradedNprotein chased

outof the cellsduringthisperiod.

Thus, in many casesthe increase in soluble

M protein observed in the mutant virus-in-fected cells at the nonpermissive temperature

wasreversedbylowering the temperature, and this protein is presumably then incorporated into the membranes of the infected cells and intoextracellularparticles.

DISCUSSION

We have utilizedtemperature-sensitive mu-tantsofVSVin anattempttodefine the

inter-relationships of the viral proteins in the proc-essesofvirionmorphogenesis. Certainaspects

of the separate pathways of maturation of the viralproteins shown schematicallyinFig.6are apparently independent of the other viral

pro-teins, suchasmigrationofthe Gproteintothe surface of the infected cell. Others, such as

binding of the M protein and nucleocapsid to

membranes, are the resultofinteractions

be-tweenviralproteins.

Effect of mutations in the glycoprotein.

Protease digestion and surface iodination ex-periments have demonstrated that the G pro-tein oftsM501(V)andts045(V) doesnot mature tothesurfaceofinfected cellsatthe nonpermis-sive temperature.Also, conversion of Gprotein

tothe sialylated

G,

form doesnotoccur under nonpermissive conditions. Cell fractionation studies have suggested that the G protein of tsM501(V) may be defective inmovementfrom the rough endoplasmic reticulum(fractions1 +

2, Fig. 4). Thus, the lack of addition of the terminal sialic acid may be a result of the ina-bility of theproteintomigratetotheproper site

inthecell forglycosylationrather thanits

ina-bility toact as a receptor for thecarbohydrate residues. However, some mutant G protein does appear in fractions 5 + 6under nonpermis-siveconditions, andit is unclearwhether this reflects imprecision of thefractionation scheme

ortruemovement into smoothmembranes (see reference 5).

The defect of the G protein isinherently in-teresting because it implies that proteins once inserted into theendoplasmic reticulum are not passively transported bymembrane movement

tothe surface of thecell.Instead, some

http://jvi.asm.org/

(9)

40S Viral

RNA

Progeny rep

Nucleocapsids

Soluble N N

Protein mRNA

RA

/ication

tsM601 G41

(Nprotein defect?)

Nucleocapsid

Soluble M mRNA / Protein transcription

I _

G mRNA

RoughEndoplasmic

Reticulum

FIG. 6. Schematic diagram illustrating the pathways of maturation of the majorstructural proteins of VSVand theproposedsiteofblock in virionassemblyforcertain temperature-sensitive mutants.

tionbetween the G protein and cellular struc-tures, possiblytransport proteins, mustoccur.

Whateverthe defectmayactually be,aportion

ofthe G protein oftsO45(V) can move to the surface of the celloncethetemperatureis

low-ered, suggestingthattheproteincanfoldtoits proper conformation and then mature

nor-mally.

Effect ofdefectsinassembly of nucleocap-sids. Two mutantsthat encodealabile N

pro-tein andsynthesize little40Sviral RNAatthe

nonpermissive temperature yielded somewhat

different results in terms of the viral protein

structures that accumulate in cells infected

with these viruses at the nonpermissive tem-perature. The maturation of the G protein to plasma membranes appeared to be normal in

bothcases, and thusthere was noevidence for

anyrole of thenucleocapsids inmaturation of the Gprotein. However, incells infected with tsM601(VI) nearly all of the M protein was

soluble, whereas in cells infected with

tsG41(IV) alarge percentage of the M protein wasinmembrane-boundstructureswith

inter-mediate density. Thesemay have been actual

intermediates of budding because at least a

portionofthesestructurescouldbechased into virionsafterashift intemperature to31°C.The

differencebetween thesetwosituationsmaybe

explained by differences in the actual muta-tions of the viruses. In cells infected with tsM601(VI)thereareprobablyveryfew nucleo-capsids in the cells at any temperature, and

under these conditions the M protein maynot

be able to bind stably to membranes. On the other hand, tsG41(IV) accumulated large amounts of 40S RNA at 31°C, and while

allowing synthesis of mRNA at 310C we must

be allowing the accumulation of significant amountsofnucleocapsids. These maybe

capa-ble ofbinding to membranes with the M

pro-tein, but incapable of budding from the cell until thetemperatureis lowered.

Incells infected with tsM502(V)noneofthe N

protein is assembledintonucleocapsids. Again,

verylittle of the M protein is bound to

mem-branes, suggesting that the M proteincannot formastablecomplexontheplasmamembrane

with only the glycoprotein. This analysis is

complicated, however, bythe fact thatwe

can-notruleoutthepossibility ofasecondmutation

inthe G proteinsuggested byits assignmentto complementationgroupV.

Effect of mutationsinthe Mprotein.The M

protein incells infected withtsM301(III)atthe nonpermissive temperature is degraded at a

ratethree-tofourfold faster than theMprotein in cells infected with wild-type virus (7). The residual undegraded tsM301(III) M protein is almostexclusivelymembranebound,and much of it bands at the density of whole virions. These structures may not be normal budding

intermediates because none of the M protein

couldbe chased into virionsuponashift-down to the permissive temperature. Furthermore, protease treatment of intact cells couldnot re-movethe membrane-boundMproteins, andno

buds wereevident on the cellsby electron

mi-ts045

M501

Golgi(?)

1157

http://jvi.asm.org/

[image:9.501.106.397.70.278.2]

(10)

KNIPE,

croscopy (unpublished data). We are presently

uncertain as to the exact identity of these M

protein-containing structures. The

tempera-ture-sensitive lesion, apparently in the M

pro-tein, had no effect on the migration of the G

protein tothecell surface, and thus there was no evidence fora role ofthe M protein in the

maturationof the G protein.

ACKNOWLEDGMENTS

We gratefully acknowledgethe technical assistance of Martin Brock.

D.K. wassupported by a National Science Foundation predoctoral fellowship during partof this work and aPublic Health Service traineeship during the remainder. D.B. is anAmerican Cancer Society researchprofessor. H.F.L. was therecipient of Public Health Service research career devel-opment award GM-50175 from the National Institute of General Medical Sciences. This work was supported by Public HealthService grants AI-08814 and AI-08388 from theNational Institute of Allergy and Infectious Diseases, American Cancer Society grant E559, and Public Health Service grant CA-12174 from the National Cancer Institute.

LITERATURE CITED

1. Flamand, A. 1969. Etude des mutants thermosensibles du virus de la stomatite vesiculaire mise en point d'untest de complementation. C.R. Acad. Sci. Paris 268:2305-2308.

2. Flamand, A. 1970. Etude genetiques du virus de la stomatite vesiculaire: clasement de mutants thermo-sensibles spontanes en groupes de complementation. J.Gen.Virol. 8:187-195.

3. Garoff, H.,and K. Simons. 1974. Locationof thespike glycoproteins in the Semliki forest virus membrane. Proc. Natl. Acad. Sci. U.S.A. 71:3988-3992.

4. Holloway, A. F., P. K. Y. Wang, and D. V. Cormack. 1970. Isolation and characterization of temperature-sensitive mutants ofvesicular stomatitis virus. Virol-ogy 42:917-926.

5. Knipe, D. M., D. Baltimore, and H. F.Lodish. 1977. Separate pathways of maturation of the major struc-tural proteins ofvesicular stomatitis virus. J. Virol. 21:1128-1139.

6. Knipe, D. M., H. F. Lodish, and D. Baltimore. 1977. Localization of two cellular forms of the vesicular stomatitisviral glycoprotein. J.Virol.21:1121-1127. 7. Knipe, D., H. F. Lodish, and D. Baltimore. 1977.

Anal-ysis of the defects of temperature-sensitive mutants of vesicular stomatitis virus: intracellular degrada-tionofspecificviralproteins. J. Virol. 21:1140-1148. 8. Lafay, F. 1974.Envelope proteins of vesicular

stomati-tisvirus:effectof temperature-sensitive mutations in complementation groups III and V. J. Virol. 14:1220-1228.

9. Ngan, J. S. C., A. F. Holloway, and D. V. Cormack. 1974.Temperature-sensitive mutants ofvesicular sto-matitis virus:comparison of the in vitro polymerase defectsof group I and group IV mutants. J. Virol. 14:765-772.

10. Pringle, C. R. 1970. Genetic characteristics of condi-tional lethal mutants of vesicular stomatitis virus induced by 5-fluorouracil, 5-azacytidine, and ethyl methane sulfate. J. Virol. 5:559-567.

11. Pringle, C. R. 1970. Theinductionand genetic charac-teristics ofconditional lethal mutants ofvesicular stomatitisvirus, p. 567-582. In R. D.Barryand B. W. J. Mahy (ed.), The biology oflarge RNA viruses. Academic Press Inc., London.

12. Rettenmier, C., R.Dumont, andD. Baltimore. 1975. Screeningprocedureforcomplementation-dependent mutantsof vesicular stomatitis virus. J.Virol. 15:41-49.

13. Unger, J. T., and M. E. Reichmann. 1973. RNA synthe-sis intemperature-sensitive mutantsof vesicular sto-matitis virus. J. Virol. 12:570-578.

http://jvi.asm.org/

Maturation of viral proteins in cells infected with temperature-sensitive mutants of vesicular stomatitis virus.