Attachment and Eclipse of Adenovirus

(1)

Attachment and

Eclipse of

Adenovirus

LENNART PHILIPSON

DivisionofCellBiology,DepartmentofMedicalMicrobiology,UppsalaUniversity,Uppsala, Sweden

Received forpublication26 June 1967

The attachment and eclipse of adenovirus have been studied with the aid of highly purified "4C-threonineand32P-labeledadenovirustype 2 in KBcells in suspen-sion cultures. Adenovirus

particles

and

infectivity

appear toattachat the same rate. The attachment rateappearstobe highly dependent on the cellconcentration and less dependent on virus concentration within the multiplicity range from 0.15 to 3 plaque-forming units per cell, probably corresponding to 4.5 to 90particles per cell. Subsequenttoattachment, 5to8% ofthe14Clabel iselutedfromthecellat a structure level, corresponding tofree hexon. The 3"2Pactivity is rapidly associated withthe cells andisconverted within20 to 30min to 65 to 85% deoxyribonuclease-susceptible material. This process isunaffectedby actinomycinandpuromycin. The deoxyribonuclease-sensitive material is, however, associated with 'IC label for an extended period after infection, and does not sediment as free deoxyribonucleic acidinsucrosegradients.Theimplications ofthesefindingson thepenetration mech-anismofanimalvirusesare discussed.

It hasbeen assumedthat theviral genomewill notbe abletoreplicateand directprotein synthe-sis within susceptible cells unless it is uncoated from the protein envelope, but since vaccinia messenger ribonucleic acid(mRNA)isapparently synthesized in the absence of protein synthesis (19, 24) and since this virusrequiresprotein syn-thesis for complete uncoating

(15, 16),

this assumptionmay beerroneous. Before

generaliza-tions are made, however, additional viruses should be

investigated

with

regard

to these properties.

Although themechanism for

uncoating

of the genome has been

partially

resolved for tailed

bacteriophages (8,

12),

thesame eventsinanimal

virus infection have only recently been studied withbiochemical

techniques (15, 16).

Itisgenerally believed thatanimal virusesare

decoatedwithin,orinclosevicinityto,

phagocyto-tic vesicles subsequent to uptake of the intact virionsby pinocytosis

(for reviews,

see 4and17). From electron

microscopic

studies, it has been concludedthat

uncoating

ofadenovirusesoccurs

at or in close

vicinity

to the nuclear membrane after

primary

pinocytosis

of thevirions

(3).

Methods for obtaining highly purified

radio-actively

labeled adenovirus have recently been

developed (9).

Since such virus

preparations

constitute a

prerequisite

for biochemical

studies

of attachment and

penetration

mechanisms,

adenovirus was selected as a model virus. An additional advantagewith adenovirus is that the structural

proteins

can be recovered ina soluble

and essentially pure form (21), which may form the basis for evaluating the biological role of animal virus proteins, if any. The present study deals with attachment and eclipse of adenovirus and the role of cellular metabolism in these events.

MATERIALS AND METHODS

Cells. KBcellsoriginally obtainedfrom

Microbio-logicalAssociates,Inc.,Bethesda, Md.,weregrownin

spinner cultures in Eagle's spinner medium (5) with

5% horse serum,

2%0

calf serum, and antibiotics.

HeLacells from the same source were grownin the

samemedium. KB cellsforplaqueassay weregrown in plastic petri dishes with Eagle's minimal essential

medium (5) with 15% calfserum and 4% tryptose

phosphatebroth.

Virus. Adenovirus type 2, the prototype strain

originally obtained from Dr. Huebner, National InstitutesofHealth, Bethesda,Md.,wasused

through-out.

Preparationofradioactivelylabeledvirus.The

proce-dure outlined by Greenand Piiia (9) was followed

withslightmodifications. KBcells inspinnercultures

wereinfectedat acelldensityof3 )K 105to 5 X 105

cells/ml withamultiplicityof 10plaque-formingunits

(PFU)/cell. Two-liter batches were processed each

time.32P-labeled viruswasproducedincellssuspended

in minimal essential medium (MEM) with citrate

instead of phosphate and 10% calfserum dialyzed

against the citrate medium. Radioactive carrier-free

orthophosphate was added to aconcentration of2.5

mcperliter of medium.Thecellswereharvestedafter

72hr, and the cellpellet wascollectedby low-speed

centrifugation (1,000 X g for 10 min). Freon

treat-ment and sedimentation on acushion of RbCl were

868

on November 11, 2019 by guest

http://jvi.asm.org/

(2)

the same as theoriginal report. The finalpurification

was achieved by layering the virus band from the

RbCl gradient on top ofpreformed CsCl gradients

(1.2 to 1.5g/ml)andcentrifugingfor 4 hrat100,000X

g. The virus band was collected and first dialyzed

against 0.05 M tris(hydroxymethyl)aminomethane

(Tris)chloride (pH 7.4) with 5 X 10-4 M

ethylenedi-aminatetraacetate (EDTA) and storedat4 C. Itwas

repeatedly observed that purified adenovirus type 2

was comparatively labile and disintegrated if stored

at -60 or-20 C, and also afterprolonged periodsat

4 C. However, when 0.25 M sucrose-10-3 M MgCl2in

0.02 M Tris chloride (pH 7.4) was used as storage

solution,thedisintegrationwasminimalforperiodsof

)4

days at 4 C.Table 1 shows thecharacteristicsof the

purified 32P-labeled virus and alsoa comparison

be-tween Tris-EDTA and sucrose buffers as storage

solutions for purified virus. The sucrose buffer was

thereforeused in thisstudy.

Virus labeled with 14C-threonine was produced in the same manner except that the cell medium

con-tained Eagle's MEM with only 0.03 mm unlabeled

threonine. Theaddition of cold threoninewas

neces-sarytoachieve maximalyieldof virus in infectedcells.

Uniformally labeled '4C-threonine with aspecific

ac-tivityof 84mc/mmole(NewEnglandNuclearCorp.,

Boston,Mass.)wasadded inaconcentration of 10Ac

per107 cells. Thesameprocedurewasused for

purifica-tion of "GC-labeled virus as described for 32P-labeled

virus. Less than1%of theradioactive labelinthevirus

was acid-solubleand

>99%o

of thelabel sedimented withvirusinfectivityinsucrosegradients.

Virusinfectivity. Infectivity ofviruswasassayed by

the plaque technique in KB-cells as previously

de-scribed(21).

Attachment oflabeled virus to cells. Attachment

oflabeledvirustocellswascarriedoutin MEM with

TABLE 1. Characteristics and spontaneous

disintegration of32P-labeledpurified

adenovirus type2in

differenit buffers

Percent- Percentage ODbat Countsper age of ofcounts

Buffer' Day 260mr/ min per 1010 counts susceptibleto

1010 PFU PFU acid- deoxyribonu-soluble cleasec

I 1 5.0 1.5 X 106 1.4 3.2

10 7.0 1.3 X 106 2.6 10.8

II 1 4.5 1.5 X 106 1.4 1.0

101 4.5 9.5 X 105 1.5 2.2

aThe composition of buffer I was 0.05 M Tris

(pH 7.4) with5 X 10-4M EDTA. Thecomposition

of buffer IIwas0.25 Msucrose-103 MMgCl2-0.02M

Tris (pH 7.4).

bOptical density (OD) withalight pathof 1 cm

corrected forlight scattering. The ratio of OD at

260m,MtoODat280mrn was 1.31.

cCounts converted to acid-soluble form by

treatmentwithdeoxyribonuclease (100

,.g/ml)

for 30minat 37C.

2% calf serum. Whendifferent media were tested for

attachment, theefficiency wassignificantly lower in all

media tested, including phosphate-buffered saline

(PBS) with or without 10-2M MgCl2 and with and

without the addition of serum. Attachment was

carried out inround-bottom tubes vigorously shaken

in a water bath at 37 Cwith a celldensity of 2.5 X 107

to 5 X 107 cells/ml. In some experiments, the cell

density was varied.

Elution of labeled virus from cells. 32p_- and 14C.

labeled virus was attached to cells for 10 to 20 min at

37 C, according to the procedure described above.

Thecellsweresubsequentlydiluted 10-fold in ice-cold

MEMandcentrifugedat500 X gfor5min; the cells

were resuspended in prewarmed medium at a cell

density of 107cells/ml andatvariousintervalssamples

were removed and the supernatant fraction was

as-sayed for radioactivity. In some experiments, the

elutedradioactivitywas analyzed by sucrose gradient

centrifugation.

Eclipse of labeledvirus.After attachmentfor5 to20

min the cells were diluted 10-fold inice-cold MEM

andcentrifugedat 500 X gfor 5 min. Thevirus-cell

complexes were again resuspended in prewarmed

MEM with 2% calf serum at acell concentration of

107cells/mlandincubated in thewaterbath at37 C.

At different time intervals, samples were taken for

analysis of the distribution of radioactivity in the

virus-cell complexes. Time-zero was assigned to the

instant ofreincubationof thecells in the water bath.

Analysis ofdistribution of radioactivity in virus-cell

complexes. Each sample of I ml of32P-labeled virus

removedfrom theincubationmixturewascentrifuged

at 500 X g for5min,and 0.3 ml of the supernatant

fluidwasassayedforradioactivity.Another 0.3 ml of

supernatantfluidwasaddedto 5mlof cold

trichloro-acetic acid (5%), and the mixture was filtered on

membrane filters (Millipore Corp.,Bedford, Mass.),

which subsequently were assayed for radioactivity.

Thecellpelletwassuspendedin2mlof0.01 M sodium

phosphate buffer (pH 7.0) with 0.01 M MgCl2 and

sonicallytreated for 15 sec;two 0.9-mlsamples were

removed for determination of cell-associated

acid-soluble and acid-insoluble material and for material

brought into solution by deoxyribonuclease

(Worth-ington Biochemical Corp., Freehold,N.J.;crystalline,

ribonuclease-free). Deoxyribonuclease treatment was

carried out with 100 ,g/ml for 30 min at 37 C.

Trichloroacetic acidwasadded inafinalconcentration

of

8.5c,,,

and all samplesprecipitatedwith

trichloro-aceticacidwerestoredat4 Cfor 18 hr beforecounting

the supernatant fluid and the sediment (the latter

taken upin 0.1 MNaOH).This procedurefor

deoxy-ribonuclease treatment rendered more than 90% of

purified 32P-labeled adenovirus deoxyribonucleic acid

(DNA) acid-soluble. The procedureis essentiallythe

same asthatdescribedby Joklik (15).

Samples of 1mlof '4C-threonine-labeledviruswere

removed from the incubation mixture, and the total

and acid-insoluble radioactivity were determined on

the supernatant fluid as described above.

Further-more,thesupernatantfluidwas plated after

centrifu-gation at 45,000 X g for 45 min, which sediments

intactvirusparticles.Thecellpellet was assayed inthe

869

on November 11, 2019 by guest

http://jvi.asm.org/

[image:2.462.32.226.424.563.2]

(3)

same way after resuspensioninPBS and sonic

treat-mentfor 15sec.

Sucrose gradient centrifugation. Sucrose gradient

centrifugation was carried out in preformed linear gradients from5to25and 20to40% sucrose in 0.01

MTris (pH 7.4)-0.1 M NaCl.Thesampleswereapplied

atthetopof the

gradients,

andcentrifugationwasfor

60 or 180 minat 100,000 X g. The sucrosegradients

weremadeonacushion ofRbCl (p = 1.40g/ml) to

preventpelletingof intactvirions.

Sonic treatment. Sonic treatment was done in an

MSE60-wultrasonicdisintegrator.

Radioactivity. Radioactivity was determined in a

Tracerlab FDI gas-flow counter equipped with a Monomol window. No corrections were made with

32P-labeledmaterial but 14Cradioactivity wascorrected

forself-absorption.

RESULTS

Attachment of32P-labeledadenovirus.32P-labeled adenovirus

containing

5 X 107PFU correspond-ingto 15,000 counts/minwasmixed witha total of2.5 X 107 KB cells

(multiplicity -2)

both at

0 and at 37 C, as described in Materials and Methods. At various intervals, samples were

re-moved,and the supernatantfluidwas

assayed

for

radioactivity

and

infectivity. Figure

1 showsthat

attachmentofadenovirusis

relatively

slowat0

C,

with an adsorbtion rate constant of1.8 x 10-3 per mlpermin.At 37

C,

85% ofthe

radioactivity

isattachedin 50min,withan

adsorption

rate

con-stant of 1.2 X 10-2per ml per min. The

radio-activity

and the

infectivity

are adsorbedatalmost

identicalrates at 37 C. Because of the

inaccuracy

of the

plaque

technique,

no

comparison

couldbe

made at 0

C,

but no

significant

reduction in

1.0

U)

0 L.

0.5

0.2

0.1

't 0

U,

:3

I..

J7-C

10 20 30 60

Time in min. FIG. 1. Attachmentof'2P-labeledadenovirus type 2 to KB cells in suspension at different temperatures.

Symbols: 0,radioactivity; 0,infectivity.

infectivity was observed in the supernatant fluid after 30minatthat temperature.

Effect of multiplicity upon attachment of 32p_ labeled virus. The effect ofvirusmultiplicityupon attachment was tested by two different proce-dures.

(i) The cell density was varied in five different concentrations from 2 X 108 to 1 X 107 cells/ml, and the dose of32P-labeledvirus was constant.

(ii) The dose of32P-labeled virus was varied in five different concentrations, with multiplicities ranging from 3 to 0.15, with the samecell con-centration of 5 X 107 cells/ml. Atdifferenttime intervals after incubation at 37 C, samples

wer4

analyzed for radioactivity in the supernatant fraction. The details of the experimental tech-nique are described in Materials and Methods. Figure 2A shows that when the cellconcentration isvaried the attachment rate is stronglydependent onmultiplicity, but whenthevirusdose is varied attachment rate is the same irrespective of mul-tiplicity (Fig. 2B). The cellconcentrationappears to be the major parameter in determining the efficiency of attachment, with higher rates of attachment at higher cell concentrations. The "percentage law" (1) formulated for virus-anti-bodyreactions thusappears to apply also to virus-cell receptor interaction, and consequently the receptors are probably in great excess at the multiplicitiesused.

Attachment of

"4C-labeled

virus. Since the cell concentration determined the rate ofadenovirus attachmentandtheratewasuninfluencedby virus multiplicity from0.15 to 3, the kinetics of attach-ment of 4C- and 32P-labeled virus could be per-formed in separate experiments. 14C- and 32p_

10 20 30 40 50 60 10 20 30 40 50 60

Time in min.

FIG.2. Effect ofmultiplicityupon attachmentof

2P-labeled adenovirus type 2. (A) Cellconcentration was

variedfrom2X 108toI X 107cells/ml,and thesame,

virusdose wasused,giving the indicated

multiplicities-(M). (B) Virus concentration was varied, giving the

indicated multiplicities (M), and the cell dose was

constantat5X 107cells/ml.

J. VIROL.

on November 11, 2019 by guest

http://jvi.asm.org/

[image:3.462.38.231.447.626.2] [image:3.462.242.432.462.587.2]

(4)

labeled viruswasmixed together withKBcellsat a celldensity of5 x 107 andamultiplicity of 1, and the mixtures were

incubated

at 37 C. At varioustime intervals, sampleswere removed and the supernatant fluid was assayed for radioac-tivity.

Figure

3 shows that 32p_ and "4C-labeled virus attachatthesame rate,althoughthe steady state may be reached somewhat earlier with the 14Clabel, with 89 and 80% of the 32p and "4G-label, respectively, attachedat 60min.

Elutionof32P-andl4C-labeledvirus. Inspite of thefact that

32p_

and'4C-labeled virus attachedto

thecellsatthesamerate, itwas

judged

necessary to study the elution of thetwo

isotopes

from the cells subsequenttoattachment. Theprocedure is outlined in Materials and Methods, andcellsata

density

of 5 X

107/ml

and

32p_

and 14C-labeled

virusat a

multiplicity

of1 wereused. After attach-ment for 15 min, the elution was followed at

varioustimes for60min.Figure3showsthatless than 1% of the 32P label is eluted from the cell within the

period studied,

but the 14C label elutes

to a

slightly

higher

rate, and6% of the labelwas

recovered in thesupernatantfluidat60min. The

fraction

eluted varied between 5 and 8% in

dif-ferent

experiments.

All of the 14C

activity

eluting

from the cellswasacid-insoluble.

Characterization of eluted 14C label. It was

tentatively assumed,

based on the results

de-scribed above, that the

slight differential

elution between 14C and 32P label may

be

ascribed to

partial

disintegration

of the

virus

at, or in close

vicinity of,

the cell surface. To establish the

approximate size

of the eluted

fraction,

11C-'b I' .0

0 0 L.

1.04

0.1

D

-.0

*03 q

It

0212

0 I.. *o1

10 20 30 40 50 60

Time in min.

FIG. 3. Attachment and elution of 32P- and 14C-labeled adenovirus type2. Symbols: @, 32Plabel; 0,

14C label; solidline, fraction unadsorbed; dashedline,

fractioneluted.

labeled hexon antigen, intact 'IC-labeled virus, and eluted 'IC label were examined in sucrose gradients asdescribed in Materialsand Methods and the radioactivity was continously recorded. Figure4shows that the'IClabel eluted from the infected cellsresides instructures sedimenting at a slower rate than that of intact virus and at the same rate aspurified hexon antigen. Hexon anti-genicity could also be demonstratedat the peak of the14Cactivity.

Eclipse ofadenovirus. Theeclipse of the virion was assayed by

following

the fate of

32P-labeled

virus after an attachment period of 10 min, as

described

in Materials and Methods. A

cell

density of5 X 107cells/ml andamultiplicity of1 wasused. Thecountsinthedifferentfractions ina

typical experiment

are shown in Table 2, and

from these datathe percentageofDNA

suscepti-bleto

deoxyribonuclease

at

different

timescanbe

calculated,

provided

thatthe

32P-label

is confined tothe DNA of thevirus. The results of Greenand

Pifia (9) supply

thebasis for

this

assumption. As

showninTable2and

Fig.

5,therateof release of a

deoxyribonuclease-susceptible

structureis

rapid,

and, within 20 min

after

the

arbitrarily

assigned

time-zero,

maximal

uncoating

of the DNA has

taken place. In

different

experiments, between 65

Intact virions

6-*Ap32)

C.4

Hexon antigen

10 E/lutedC 1 -l/abel

c

5.

10 30 50 70 100

Per centofgradientvo/urne

FIG.4. Characterizationof the eluted14Clabel from

adenovirus-infected KB-cells. Linear sucrosegradients

from20to40%sucrosecentrifuged for 1 hr at100,000' X g.Fromtoptobottom,intact32P-labeled adenovirus type 2,purified "4C-labeled hexon antigen, and the

eluted14Clabelfrominfected KB-cells.

on November 11, 2019 by guest

http://jvi.asm.org/

[image:4.462.236.417.366.602.2] [image:4.462.31.222.438.627.2]

(5)

TABLE 2. Fate ofadenovirus type 2 labeled with3SP

Cell-associated counts In medium

PercentDNA

(B)

~~~~~~~~~~~~DNA

available susceptible

Time (min) (cid-insob) e foreclipse, to

deoxyri-(Aci-noul

sensitiveto (C) (D) (E) A+C+ E bonuclease,

Acidinsouble deoxyri- Acid-soluble Acid-insoluble Acid-soluble B/F X100 bonuclease

5 1,854 576 244 323 0 2,098 27

10 1,810 852 172 322 0 1,982 43

15 1,997 1,042 150 372 0 2,147 48

20 1,900 1,360 193 391 0 2,093 65

30 1,723 1,433 319 389 0 2,111 68

45 1,657 1,387 390 317 0 2,149 65

and 85 %of virusDNA wassusceptibleto deoxy-ribonuclease. A minor fraction of the

32P-label,

not exceeding 10% of the DNA in 4 to 6 hr, is subsequently converted to acid-soluble material insidethecell.

Multiplicity of infection andrelease of a deoxy-ribonuclease-susceptible structure. The influence

of

multiplicity

upondeoxyribonuclease

sensitivity

wastested bothby addingastandardvirusdoseto cells ofvaryingdensities andby varyingthevirus

dose at the same cell

density

level.

32P-labeled

virus was attached for 10 min at 37 C, and the percentage of 32P label

susceptible

to

deoxyribo-nuclease was determined after an additional

period

of 30 min at 37

C,

as described above.

Figure 6 shows that less

deoxyribonuclease-sensitive material isreleased athighcelldensities (A) and that the percentage ofthis material de-creases at

high multiplicities

of virus

(B).

Effect

of metabolic inhibitors upon the

appear-anceof deoxyribonuclease sensitivity. The

partici-pation

of cellular metabolism in the

eclipse

was

evaluated

by studying

the

efficiency

of

uncoating

in thepresenceof

actinomycin

Dand

puromycin,

which inhibit

DNA-dependent

RNA

synthesis

and

protein synthesis, respectively.

The dose was

5 mg of

actinomycin

perml and 50

,ug

of

puro-mycin

per ml. These concentrations were shown

toinhibitRNA

andprotein synthesis,

respectively,

to 1% ofactivities incontrolKB cells within 30 min after addition of the

inhibitors,

when in-organic 32pwas

incorporated

into RNA or 14C-threonineinto

proteins

of cells in

spinner

cultures. KB and HeLacellsweretherefore

pretreated

for varyingperiodsat37Cinspinner mediumin the presence and absence of the inhibitors at these concentrations.Viruswas

subsequently

added

by

resuspendingthecells in attachmentmediumwith andwithout inhibitors butwith32P-labeledvirus. After15minofattachment,thecellswerewashed

and

again

incubated at 37 C in the presence or

absence ofinhibitorsfor30minat 37

C,

and the percentage of

deoxyribonuclease-sensitive

ma-qJ

qj

U)

qj Q

0..

70 T

. 60t

50+

40+

30+

20 t

10

t

10 20 30 40 50

[image:5.462.234.442.101.456.2]

Time in

min.

FIG. 5. Kinetics of conversion of intact adenovirus

intoadeoxyribonuclease sensitive structure after

virus-cellinteraction.

terial was determined. Figure 7 shows that ir-respective of the time for pretreatment actino-mycin does not

affect

the transfer of virus into deoxyribonuclease-sensitive material. The pres-ence of puromycin appeared to slightly increase thecapacityofthecells to eclipse virus, especially ifit waspresent for periods of 2 to 4 hr prior to virus attachment. No difference between HeLa and KB cells could be discerned with regard to the effect ofpuromycin.

Characterization of the cell-associated32p_ and

"4C-labeled

virus.The attached virus thusappears to be rapidly eclipsed into a deoxyribonuclease-sensitiveform, and thisprocessdoes notappearto require RNA or protein synthesis. Since a core structure ofadenoviruses has been observed by electronmicroscopy (6, 14), it wasjudged

on November 11, 2019 by guest

http://jvi.asm.org/

(6)

L.

'

b 60

|, 50

o 4

.--0 t 40~ <. 30 q:

A

1o

B

0.1 1 10 0.1 1 10

[image:6.462.32.222.57.228.2]

Multip/icily

FIG. 6. Effect ofmultiplicityuponthe conversion of

intact adenovirus into a deoxyribonuclease-sensitive

structure. (A) The cell dose was varied to give the

indicatedmultiplicities. (B) The virus dose was varied

togivethe indicatedmultiplicities.

Methods, and the fractions were monitored by radioactivity and immunodiffusion. Pure 1C-labeled hexon antigen preparedasdescribed else-where (21),

extracted32P-labeled

DNA (10), and intact'4C-labeled virus were used as standards to measure thesedimentationrateofthe productsin sucrose gradients. The standard preparations wereaddedto sonically treated normal KB cells at the same cell density as in the

experimental

samples. Figure 8 shows that themajor peak of

32P-activity

recovered fromcell-associated

adeno-virus preparations is confined to amaterial sedi-menting midway between viral DNA and virus infectivity. Only about

10%

of the 32P

activity

sedimentsatthesame rate asviralDNA,but very littleasintact virions.The 32P

activity

ofthe inter-mediate peak is

susceptible

to

deoxyribonuclease.

When

cell-associated

14Clabelwas

analyzed,

it

was first demonstrated that radioactivity was

qJ1.

.U

E

qj 100

qj

80

0

s 60 c 40

q 2

Q.

0

-..°

^

I.

e- --0

f

I.

4 3 2 1 Control

Timeofpretreatment

[image:6.462.239.422.284.592.2]

in hours

FIG. 7. Effect ofmetabolic inhibitors on the

con-version ofintact adenovirus into a

deoxyribonuclease-susceptiblematerial. KB cells werepretreated for the

indicatedtimeswithactinomycin(A\)at aconcentration

of5,ig/mlorpuromycin (0) at aconcentrationof50

,tg/ml.HeLacells(0) werealso tested with puromycin atthesame concentration.

tanttoattemptto characterize the

cell-associated

product of adenovirus. The cell-associated "2P labelwasfirstexamined.

32P-labeledvirus was attachedata

multiplicity

of 1 andacell

density

of5 X 107cells for 15 min at 37C, and, afteran

eclipse

period

of60min at 37C,thecellswerewashed

twice,

resuspended in

0.5 ml, and

sonically

treated for 15 sec. This

material was fractionated by sucrose gradient

centrifugation

as described in Materials and

10 3 50 70 100

Percent of gradient volume

FIG.8. Characterization of thze cell-associated 32P

and14Clabels fromadenovirus-infectedKBcells. Linear

sucrose gradients from 5 to

25%c

sucrose centrifuged for3hr at 100,000X g. Thepatterns

shIowIz

are, from

top to bottom: 14C-labeled intact virions; 32P-labeled

adenovirus DNA; 14C-labeled hexon

antigent;

cell-associated14Clabel and cell-associated 32p label.

on November 11, 2019 by guest

http://jvi.asm.org/

[image:6.462.32.225.290.481.2]

(7)

PHILIPSON

acid-insoluble aftereclipse periods of 30 min to 4 hr. On the other hand, the 'IC label rapidly converted to a structure not

sedimenting

at 45,000 x gfor 45 min. Intact virions are, how-ever,

sedimented

atthis

speed.

Therateof appear-ance of this

lighter

structure was similar to the rate ofappearance ofa

deoxyribonuclease-sensitive

material, previously shown in

Fig.

5. When the

cell-associated

'4C-labeled

activity

was

analyzed

in sucrose gradients, about 45% of the activity wasrecovered inastructure

sedimenting midway

betweenviralDNAandintact

virions,

andatthe same rate as the

major

peak of cell-associated

39P

activity,

as shown in

Fig.

7. The

remaining

14C

activity

wasrecovered inmaterial

sedimenting

at the same rate as hexon

antigen.

A similar pattern wasobservedaslateas4 hrafter

infection;

i.e., most of the

cell-associated

32P-activity

was recovered in the structure with an

intermediate

sedimentation

rate and 40 to

50%01

of the 14C

activity

wasalso recovered in thisfraction.

DISCUSSION

The

following

model of adenovirus attachment

and

eclipse

has been

adopted

from the presented

results.

Adenovirus

particles

are

rapidly

attached to cells(80to90% in60

min),

and

partof

theprotein coatappearsto be releasedat,orinclose

vicinity

of,

the

cell

surface,

exposing

a

nucleoprotein

structureinside the

cell,

thenucleic acid of which is sensitive to

deoxyribonuclease (65

to 85%

efficiency).

Thefurther breakdown of this

struc-ture into DNA and

protein

cannot be

detected

biochemically

at4 hrafter infection. Thepresence

of14Clabel inthesizerangeof the hexon

antigen

inthecellmediumand thewashed

pellets,

aswell as the intermediate sedimentation rate

between

virus and free DNA, for the

major

part ofthe

cell-associated

32plabel andhalf of the14C

label,

support this theory.

Possibly

such a structure could

correspond

totheadenoviruscore,

observed

in the

electron

microscope (6, 14),

but

additional

evidence

concerning

thenature andfinal fate of thisstructureis

required.

Alternatively,

the inter-mediate structure may be released DNA with attached hexon

antigen,

since this

protein

has a

highDNA

affinity

(unpublished

data).

Considering

the data

furnishing

the basis for

thismodel,it was first shown thatadenovirustype 2infectivity

rapidly

attachedtoKBcells and that purified radioactive virusattachedatthesamerate

asinfectivity,in

spite

ofthe fact that the

efficiency

of the plaqueassay hasbeenestimated tobe3% ofthe number of

particles (11).

The ineffective particlesthus appeartoattach anduncoatto the same extent asthose

initiating

infection,

and the

highratio ofparticlesto PFU maybedue to in-efficiencyatsubsequentstepsofthemultiplication cycle.This is inagreement with thefindings in the vaccinia virus system (15, 16), butin contrastto the poliovirus system (18). In the latter system, however, the ratio of PFUtoparticles isaslowas 0.003.

Itis of interest that the attachment of adeno-virus particles to cells at low multiplicities appears to follow some of the characteristics established for virus-antibody union. Thus, the attachment rate is primarily dependent on cell concentration, andthe"percentage law" (1) ap-pears to applywhen the samecellconcentration is used but the virus dose is varied (Fig. 2). Fazekas de St. Groth (7) made the same pre-diction for

virus-erythrocyte

complexes.

The initial phase in adenovirus eclipse, the exposure ofviralDNA to

deoxyribonuclease,

isa rapid process with regard to the long eclipse phase in the adenovirus system. A preliminary note by Lawrence and Ginsberg (Federation Proc. 24:379, 1965) described similar observa-tions. In contrast to the vaccinia virus system (15,16), cellular protein and nucleic acidsynthesis do notappear to be required in thisprocess; on the contrary, inhibition of protein synthesis in-creases slightly, butsignificantly, theappearance

of the

deoxyribonuclease-susceptible

material

(Fig. 7).

The core structure in adenoviruses proposed from electron microscopic studies (6, 14) may possibly correspond to the cell-associated struc-ture containing both 14C and

32p

label (Fig. 8).

The

prolonged

eclipse phase of adenovirusmayin

thatcasebe dueto theuncoating of this material ifadditional uncoating is necessary for transcrip-tionortranslationfrom the viralgenome.Recent findings in the vaccinia virus system (19, 25) indicate thatacoatedgenome canbe transcribed. No enhancement of

eclipse

could be observed

at

higher

multiplicities;

on the contrary,

eclipse

was

relatively

less efficientat

higher multiplicities

and at

higher

cell concentrations. With

vaccinia

virus (15, 16), DNA uncoating was in contrast morerapid at

higher

multiplicities, which,

how-ever, maybeinfluenced

by

theapparentinduction mechanisminvolvedinuncoating inthat system. Theimplicationsofthefurnished resultsforthe mechanismofviruspenetrationand

eclipse

merit a comment.

Generally,

animal viruses are pro-posedtoenterthe cellby

pinocytosis

followedby

disintegration

within

pinocytotic

vesicles

(4,

17).

The present results are, however, not

entirely

compatible withtheelectron

microscopic

dataon

adenovirus

penetration

and

eclipse (3).

Thus, the rapid disintegration of 65 to 85% of the virions J. VIROL.

on November 11, 2019 by guest

http://jvi.asm.org/

(8)

observed in this study is in contrast to the slow penetration and accumulation of intact particles within the cell, observed by electron microscopy. Differences in methods of purification of the virions maypossiblyexplainthediscrepancy.

The pinocytosis concept for virus penetration should, furthermore, be considered in the light of Cohn's studies on the effect of antimetabolites upon pinocytosis of macrophages (2), where actinomycin andpuromycin in concentrations of 0.01 and 0.1,ug/ml,respectively, inhibit pinocyto-sis toaround

107%

of the controlactivity. Inthe present study, actinomycin and puromycin at 5 and 50 ,ug/ml, respectively, failed to inhibit adenovirus disintegration. In addition, studies on picornavirus eclipse have furnished evidencethat the virion is modified or even uncoated in the presenceofcell membrane preparations (13, 22). Furtherstudiesontheuncoatingofpoliovirus by intact cellswashampered bythe elution ofvirus from the cells (18), probably complexed with receptor structures (23), andbytherapid

break-down of the parental genome into acid-soluble

form(18).

Thus, it is stillpossiblethat animal viruseswith cubicsymmetry willnotbeengulfed bythe

viro-pexis

mechanism, incontrast tothe

lipid-contain-ing viruses.

ACKNOWLEDGMENTS

Thisinvestigationwassupported bygrantsfrom the

DamonRunyonMemorialFund,theSwedishMedical

ResearchCouncil,andtheSwedishCancerSociety.

LITERATURE CIED

1. ANDREWES, C. H.,AND W. J. ELFORD 1933.

Ob-servations on antiphage sera. Brit. J. Exptl.

Pathol. 14:367-383.

2. COHN,Z.A.1966. Theregulationofpinocytosisin

mousemacrophages. I.Metabolicrequirements

as defined by the use of inhibitors. J. Exptl.

Med. 124:557-571.

3. DALES, S. 1962. An electronmicroscope study of

the early association between twomammalian

viruses and their hosts. J. Cell Biol.

13:303-322.

4. DALES, S. 1965. Penetration of animalvirusesinto

cells. Progr. Med. Virol. 7:1-43.

5. EAGLE, H. 1959. Aminoacid metabolism in

mam-malian cell cultures. Science130:432-437.

6. EPSTEIN, M. A.,S. J. HOLT,AND A. K. POWELL.

1960. The fine structure and composition of

type S adenovirus; an integrated electron

mi-croscopical and cytochemical study. Brit. J.

Exptl.Pathol. 41:567-576.

7. FAZEKAS DE ST.GROTH,S. 1962.The

neutraliza-tion ofviruses. Advan. VirusRes. 9:1-125.

8. GAREN,A.,ANDL. M. KOZLOFF. 1959.The

initia-tion of bacteriophage infection, p 203-236.

In F. M. Burnet and W. M. Stanley [ed.],The

viruses, vol. 2,AcademicPress, Inc., New York.

9. GREEN, M., AND M. PINA. 1963. Biochemical

studiesonadenovirusmultiplication.IV.

Isola-tion, purification and chemical analysis of adenovirus. Virology 20:199-207.

10. GREEN, M., AND M. PINA. 1964. Biochemical

studiesonadenovirus multiplication. VI.

Prop-erties of highly purified tumorigenic human

adenoviruses and their DNA's. Proc. Natl. Acad. Sci. U.S. 51:1251-1259.

11. GREEN, M., M. PINA, AND R. C. KIMES. 1967. Biochemical studies on adenovirus

multiplica-tion. XII. Plaquing efficiencies of purified

humanadenoviruses.Virology 31:562-565.

12. HERSHEY, A. D., AND M. CHASE. 1952. Independ-entfunctionsof viralproteinand nucleic acid in

growth ofbacteriophage. J. Gen. Physiol. 36:

39-56.

13. HOLLAND, J. J. 1962. Early stagesof enterovirus

infection. Cold Spring Harbor Symp. Quant.

Biol.27:101-111.

14. HORNE, R. W. 1963. Architectural symmetry in

viruses and their components. Perspectives

Virol. 3:43-57.

15. JoKLIK, W. K. 1964. Theintracellular uncoating ofpoxvirus DNA. I. The fate ofradioactively

labeled rabbit poxvirus. J. Mol. Biol. 8:263-276.

16. JOKLIK, W. K. 1964. The intracellular uncoating ofpoxvirus DNA. II. The molecular basis of theuncoatingprocess. J.Mol. Biol.8:277-288.

17. JOKLIK, W. K. 1965. The molecular basis of the

viraleclipsephase. Progr.Med.Virol. 7:44-96.

18. JOKLIK, W. K., AND J. E. DARNELL. 1961.

Adsorb-tion and early fate ofpurified poliovirus.

Virol-ogy13:439-A47.

19. KATES, J. R., AND B. R. MCAUSLAN. 1967.

Mes-senger RNA synthesis by a "coated" viral

genome. Proc. Natl. Acad. Sci. U.S.

57:314-320.

20. PETTERSSON, U., L. PHILIPSON,AND S. HOGLUND.

1967. Structural proteins of adenoviruses. I.

Purificationand characterizationofadenovirus

type 2hexonantigen. Virology,inpress.

21. PHILIPSON, L. 1961. Adenovirus assay by the

fluorescent cell-counting procedure. Virology

15:263-268.

22. PHILLPSON, L., AND M. LIND. 1964. Enterovirus eclipsein acell free system. Virology 23:322-332.

23. PHILIPSON, L., AND S. BENGTSSON. 1962.

Inter-action of enteroviruses with receptors from

erythrocytes and host cells. Virology

18:457-469.

24. WOODSON, B. 1967. Vaccinia mRNA synthesis

under conditions which prevent uncoating.

Biochem.Biophys. Res. Commun. 27:169-175.