Electron Microscopy of Cells Infected with Semliki Forest Virus Temperature-Sensitive Mutants: Correlation of Ultrastructural and Physiological Observations

JOURNALOFVIROLOGY,May 1970,p.632-638

Copyright( 1970 AmericanSocietyforMicrobiology PrintedVol.in5,U.S.A.No.5

Electron

Microscopy

of

Cells

Infected

with

Semliki

Forest Virus

Temperature-Sensitive

Mutants:

Correlation

of

Ultrastructural

and

Physiological

Observations

K. B. TANI

DepartmentofMicrobiology, John Curtin School of MedicalResearch, Auistralian

NationalUniversity, Canberra, Australia

Received for publication 16January 1970

Cellsinfected at thepermissivetemperature with threetemperature-sensitive mu-tants of Semliki Forest virus were not significantly different in appearance from cellsinfected,ateither thepermissiveornonpermissive temperature,withwild-type virus. Virus particles, nucleocapsids, spherules, and tubules were seen in the cyto-plasm. But replicationof the mutants was inhibited incellsinfected at the

nonper-missive temperature. This was evidenced by the absence ofvirus particles and nucleocapsids (exceptinonecase) and the absence or limitedproduction of spher-ules and tubspher-ules. These observationsarediscussedwithreference to the physiologi-caldefects of the mutants.

Temperature-sensitive (ts) mutants have been

isolated from Semliki Forest virus (SFV) and

characterized indetail (5;Tan,unpublisheddata). These mutantsreplicated normally in cells

incu-batedat 30 C (permissive

temperature)

butwere

blockedatvarious stages ofthe replication cycle

at 38.5 C (nonpermissive temperature). Wild-type SFV replicated normally at both 30 and

38.5 C.

Cells infected with SFV have been studied in detail in theelectronmicroscope (1, 2, 3).Peculiar totheinfected cellswerevacuoles surroundedby

viral nucleocapsids, aggregates of viral

nucleo-capsids, virus particles, and spherular and

tubu-lar structures, all of which were seen in the

cytoplasm. Virus particles were formed by

nu-cleocapsids budding at cellular membranes to

acquire an envelope. The significance of the tubular and spherularstructuresisnot known.

As thephysiologicaldefects of the mutants are known (5), itwastherefore ofinterestto examine in theelectron microscope and then to compare the ultrastructure of cellsinfected withwild-type virusorwithmutants. This paper reportsstudies

with mutants which, at the nonpermissive

tem-perature, are blocked in viral functions subse-quent to viralribonucleic acid (RNA) synthesis

(5).

1Present address: Department of Microbiology, Roche

Insti-tuteof MolecularBiology,Nutley, N.J.07110.

MATERIALS AND METHODS

Cells.Thesamepreparation of primary chick

em-bryo (CE) cells, grown inmonolayers,wasusedin all

oftheexperiments. The preparation and maintenance

of CEmonolayershave been described (5).

Virus. The infection of cells with and growth of

SFV and ts mutants were asdescribedby Tanetal.

(5), except that actinomycin D was omitted in the

present study and infected cells were sampled at 14

hrafter infection.

Wild-type virus and three RNA+ mutants were investigated. These mutants, ts-15, ts-18, and ts-21, at

38.5 Csynthesize 13, 24, and 14%, respectively, of the

wild-type yieldof RNA (5; Tan, unpublished data).

All threemutants make viral membrane protein but

only ts-15 makes nucleocapsids as well. Thus, the

temperature-sensitive defectof both ts-18 and ts-21 is

in theproduction of viral nucleocapsids and that of

ts-15maybematuration of virus particles (5).

Electron microscopy. Cells were scraped off the

glass,pelleted, fixed with 1% gluteraldehyde [in

phos-phate-buffered saline (PBS)] for5to 10min at room

temperature, andpostfixed with1% oxmium tetroxide

(in PBS) for 60 to 90 minat 4 C.After washing with

PBS containing 10% sucrose, the cell pellets were dehydrated in acetone. All cell pellets were stained,

duringthe dehydration stage, in asaturated solution

ofuranyl acetatein70% acetone. After dehydration

in 100% acetone, the cell pellets were embedded in

arladite (as instructed, "Durcupan," ACM, Fluka).

ThinsectionscutwithaReichert microtome with glass

knivesweremounteddirectly on 400-mesh grids. The

sectionswerethenstained with0.2% lead citrate in 0.1

632

on November 11, 2019 by guest

http://jvi.asm.org/

CELLS INFECTED WITH ARBOVIRUS ts-MUTANTS

N NaOHfor1 min, washed, and examined in a Philips

EM 200electron microscope.

RESULTS

Cells infected at30 C. Cells infected for 14 hr at30 Cwithwild-typevirusorwithmutants were not significantly different in appearance [see Acheson and Tamm(1) andErlandsonetal. (2)].

Infectious virus was still being made at this late time after infectionat30 C (5).

Several sections ofeach samplewereexamined,

and representative electronmicrographs ofthese are shown in the accompanying figures. Infected cells characteristically contained virus-associated vacuoles, viral nucleocapsids, virus particles, tubules, andspherules.

Virusparticles andnucleocapsids. Infectedcells werehighly vacuolated compared with uninfected

cells. Many of thevacuoles wereassociated with viralstructures,

being

surrounded

by

viral nucleo-capsids and frequently also containing virus

particles (Fig. 1, 2a, 3a, 4a).

Nucleocapsids often occurred in aggregates but small numbers of

single nucleocapsids

were

also observed scattered inthe cytoplasm (Fig. 1,

2a, 3a,4a). Virus

particles

wereformedby nucleo-capsidsbuddingatthe

plasmalemma

(not shown;

1, 2) or at the vacuolar membrane (Fig. 2a) to

acquirean

envelope.

Spherules and tubules. Spherules are saclike

membranous

atructures

containing an electron-dense particle in the center (Fig. 4a). They are seen only in infected cells, appearing soon after infection (1, 3). Cells infected at 30 C with wild-type virus or with mutants contained spherules.

inthe cytoplasm (Fig. 1) and on thecell surface (Fig. 4a), frequently attached to extensive areas onthe cell surface.

Tubules appeared late in infection, and in in-tact cells they were found in virus-associated

vacuoles (Fig. 1, 2a, 3a; references 1, 2, 4).

Tu-bules appeared to originate from the vacuolar membrane (Fig. 5a). Virus particles were often

attached to the ends of tubules (Fig. 3a), and tubulesmaycontainmorethanone

nucleocapsidc

arrangedin a linear order (Fig. 5a). Like spher-ules, tubuleswereabsent inuninfectedcells.

Cellsinfected at38.5 C. Only the cellsinfected

at 38.5C with wild-type virus (not shown)

re-sembled cells infected atthepermissive tempera-ture.

In cells infected at 38.5 C with either ts-15 (Fig. 2b), ts-18 (Fig. 3b), or ts-21 (Fig.

4b),

virus particles and budding virus particles were

absent, tubules were absent, and

nucleocapsids,

notinsuch massiveaggregatesas werepresentin

UK...

FIG. 1. Portionofacellinfected with wild-typevirus at 30 C.Thecytoplasm is

hightly

vacuolated. Some vacuoles

aresurroundedbyviralnucleocapsids (A),and others are surrounded by nucleocapsids and contain virus particles

andtubules(B). Spherules (arrows)arepresentinthecytoplasm. X34,000.

633.

VOL.5, 1970

on November 11, 2019 by guest

http://jvi.asm.org/

[image:2.486.46.441.360.606.2]

TAN

r~~~~~~~~~~~~~r

A AAt

~~j~\~~b44

*

I,.

.4

sr

..t

tAi 1,V I.,,

:..-".

r\

Wr, C', i.r t Wg ,r

if l.r..

[image:3.486.57.448.72.569.2]

.)

FIG. 2. Portionsof cells infectedwith ts-15. (a) Cellinfectedat30Cshowingacollectionofvirus-associated

vacuoles andaggregates of nucleocapsids in the cytoplasm. In one vacuole, virusparticles areformed by

bud-ding (arrows). X39,000. (b) Cellinfectedat38.5C. Nucleocapsids (arrows) arescattered in the cytoplasm.One

vacuoleislinedonits inneredge by spherules. XSI ,000.

634

J. VIROL.

- .' ...

.

.-1. I,

6."

on November 11, 2019 by guest

http://jvi.asm.org/

VOL. 5,1970 CELLS INFECTED WITH ARBOVIRUS ts-MUTANTS 635

t.%$- j + j *-!'

4.i

JO

.

4\8*

*--'*"V

.$*v& X

a F.jQ~ ' i v

,'' * .t

.4A

w-1,1 t

;:.i.,-.46rel

[image:4.486.47.440.79.576.2]

A~

FIG. 3. Portionsofcellsinfectedwith ts-18.(a) Cellinfectedat30C. Intwovacuoles,virusparticlesareattached

totheendsoftubules (arrows).Somevacuoles(A, B) containgroupsof nucleocapsidsandvirusparticles,someof

whicharebudding. Such vacuolesareprobably invaginated[Erlandsonetal. (2)]. X 64,000. (b) Portionsoftwo

cellsinfectedat38.5 C.Numerousspherulesarepresentbetween thetwocells.X35,000.

.t

2i

4s.

oR:.Y

Jr.

on November 11, 2019 by guest

http://jvi.asm.org/

636 TAN J. VIROL.

£4

-'~~~~~~~~~~~~~~~~~~~~~~~~

V' : * rS? i 'aS**o

r vtS -tigt

;

tw>,Z**f3f-t

-~ A lg;t*44 ;r+

ft'¢iev

Y.t;

\tbj*44*ntt

"*t

'4

A.

r!

@%>

.0';

-

} 4t s S A n, >

k~ ~ ~ ~ * T k

a tt

;gyt>-*

i41

1#

4~~~~~~~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~.~4~~~~~s444

V~~~~~~~Vc

i':;~~~~~~~~~~~~~~~~~5

:N'f;

4~~~~~~~~~~~~~

44$-~~~~~~~~~~~~~4

I.. -hr~~~~~~~~~~~~~.

*~~~~

4*4

WsP ?,*;4 (* I

FIG. 4. Portions of cellsinfectedwithts-21. (a) Cell infected at 30C. Nucleocapsids areattachedto the cyto-plasmic side ofthteplasmalemma (short arrow), and spherules are attached to the outer side (long arrow). X47,000. (b) Cell infectedat 38.5 C. Subcellular organelles are well preserved.

Thte

significance of rare extracellular virus particles is not known.

X(33,000.

on November 11, 2019 by guest

http://jvi.asm.org/

[image:5.486.57.447.71.573.2]

CELLS INFECTED WITH ARBOVIRUS ts-MUTANTS

[image:6.486.44.435.66.275.2]

b

FIG. 5. Tubules. (a) Longitudinalsectionoftubules attachedtovacuolar membrane. (b) Transversesection of

tubules(arrow). (c) Tubules containingnucleocapsids. Scale line= 100nm.

cells infected at 30C, were seen only in cells infectedwith ts-15.

Spherules, fewer than incells infectedat30C, were observed in cells infected at 38.5 C with

ts-15 (Fig. 2b) or ts-18 (Fig. 3b). Spheruleswere

absent in all sections of cells infected at 38.5 C

with ts-21 (Fig. 4b).

DISCUSSION

Correlation of ultrastructural and physiological

observations. Wild-type SFV multiplies equally

well at both 30 and 38.5C, producing normal

amounts ofviral nucleocapsids, membrane pro-tein, and virus particles, but tsmutants produce normal amounts of viral components only at

30C (5). These observations were confirmed in

thepresent study of infected cellsinthe electron

microscope.

At 38.5C, ts-15 made nucleocapsids but not

complete virusparticles, whereas both ts-18 and

ts-21 appeared defective in the production of

nucleocapsids, as investigated with the electron microscope.Thesefindingsareincomplete

agree-mentwith the results obtained withphysiological

tests (5). In those tests, it wasconcluded that at

38.5 C all three mutants made viral membrane

proteinbutonlyts-15 madenucleocapsidsaswell. Morphologically, the cellular membrane and

viral membrane (or envelope) are similar in

ap-pearance, except for the presence of spikes on

the latter. Inthe maturation of the virusparticle,

spikeswere seen onlyonbuddingvirus particles

(1; unpublisheddata). Thus it was not possible, in thepresent study, to idenfity in cells infected with the mutants at 38.5 C cellular membranes

which were presumably altered biochemically by the insertion of viral membrane protein in them (5).

Spherules and tubules. One difference between the cells infected with mutants at 38.5 C and at 30C is the reduced number of (with ts-15 and

ts-18) or the absence of(with ts-21) spherulesat

38.5C. Mutantsts-15, ts-18,and ts-21 makeviral

RNA at 38.5 C (5), but make few or no spherules

under similar conditions. This suggests that spherules are notassociated with viralRNA

syn-thesis. Spherules werealways present in cells in-fectedat30C,atwhich temperaturenormal viral

replication occurred. Thus, spherulesmay be

re-lated to viral replication other than viral RNA

synthesis and their nature and function remain

tobeelucidated.

Tubuleswerenotdetected in cells infectedwith

mutants at 38.5C. Thus, it seems that tubule production is associated with the ability of the

virus to complete thenormal growth cycle. My observations that (i) virus particles and tubules

were of a similar diameter (48 and 42 nm,

re-spectively), (ii)tubuleswereattached tothe mem-brane ofvirus-associatedvacuole, (iii)virus

parti-cleswere found at the ends oftubules, and (iv)

some tubules contained more than one

nucleo-capsid ledto theinterpretation thata

nucleocap-sidbudding into a vacuolewasnot "nipped off"

to forma virus particle; instead, the viral mem-637

VOL.5, 1970

.-owl 7

,it

on November 11, 2019 by guest

http://jvi.asm.org/

638 TAN

braneprotein continued tobe made,resulting in

theformation ofa tubule. If othernucleocapsids

happen to bud at the same spot (not unlikely), then the resulting tubule will contain more than one nucleocapsid. Therefore I conclude that tu-bulesare viral membrane protein resulting from

abnormal maturation of virus particles late in

infection. At such a time, the resources for nor-mal maturation of virus particles may be

re-stricted.

The defects of conditional lethal mutants of animalviruseshave beencharacterizedon physio-logical, biochemical, and genetic grounds.

De-tailed electron microscopic studies of cells in-fectedwithmutantshavenotbeenreported.Such

studies could providenewandvaluable

informa-tion on the defects of the mutants and thus the

replicationofwild-typevirus.

J. VIROL.

ACKNOWLEDGMENTS

Iamindebted toG. I.Schoefl of the Department of Experi-mental Pathology and M. Taylor and J. Preston of theElectron Microscope Unit, JohnCurtinSchoolof Medical Research, Can-berra, Australia, for expert instructions in the technique of elec-tronmicroscopy.

LITERATURECITED

1. Acheson, N. H., and I. Tamm. 1967. Replication of Semliki Forest virus: an electron microscopic study.Virology 32:128-143.

2. Erlandson, R. A., V. I. Babcock, C. M. Southam, R. B. Bailey, and F.H.Shipkey. 1967.Semliki Forest virus inHEp-2 cell cultures. J.Virol. 1:996-1009.

3.Grimley, P. M., I. K. Berezesky, and R. M. Friedman. 1968. Cytoplasmic structures associated with an arbovirus infec-tion: loci of viral ribonucleic acid synthesis. J. Virol. 2:1326-1338.

4.McGee-Russell, S. M., and G. Gosztonyi. 1967. Assemblyof Semliki Forestvirusin brain. Nature(London) 214:1204-1206.

5.Tan, K. B., J. F.Sambrook, and A. J. D. Bellett. 1969. Semliki Forest virus temperature-sensitive mutants: isolation and characterization.Virology 38:427-439.

on November 11, 2019 by guest

http://jvi.asm.org/