Mouse peritoneal cells confer an antiviral state on mouse cell monolayers: role of interferon.

Copyright© 1986,American Societyfor Microbiology

Mouse

Peritoneal Cells Confer

an

Antiviral State

on

Mouse

Cell

Monolayers: Role

of

Interferon

ENRICOPROIETTI,1* SANDRAGESSANI,1 FILIPPO

BELARDELLI,'

AND IONGRESSER2

Departmentof Virology, Istituto Superiore di Sanita', Rome,

Italy,'

andLaboratory of Viral Oncology,Institut de Recherches Scientifiques surleCancer, Villejuif, France2

Received4March1985/Accepted 3June 1985

Vesicular stomatitis virus and encephalomyocarditis virus do not multiply in the majority of peritoneal macrophages freshly explanted from 4- to 8-week-old male or female mice. However, when peritoneal macrophages werecultivatedin vitro for 3 to 5 days,these cells became permissive for both viruses. The loss of antiviral state in "aged" macrophages paralleled a significant decrease in the intracellular levels of

(2'-5')oligo-adenylate synthetase activity. Although biologically active interferon was not detected in the

nutrientmediumofmacrophage cultures, freshly harvested

peritonea!

cells could confer an antiviral state on monolayer cultures of mouse cells (aged macrophages, embryonic fibroblasts, and L cells) but not on heterologous chicken embryo, rabbit kidney, or human cells infected with vesicular stomatitis virus or encephalomyocarditis virus. The conferred antiviral state required at least 7 h to develop in target cells and was

totallyinhibited by the presence of antibody to mouse interferon

a/1

but not tointerferonyin thecocultures.

Heterologous guinea pig and rabbit peritoneal cells could not transfer an antiviral state to targetmouse cells. Donorperitoneal cells from mice preinjected with antibody to interferon

a$P

couldnot transfer an antiviral statetotarget mouse cells. This ensemble of results indicating that freshly harvested peritoneal cells transfer interferon(which is responsible forinducing an antiviral state in susceptible mouse target cells)

adds

further experimental evidence that interferon is spontaneously expressed in normal mice and plays an important role in maintaining some host cells in anantiviral state.

Theresistance of restingmouseperitonealmacrophages to

vesicular stomatitis virus (VSV) and encephalomyocarditis

virus (EMCV) can be abolished by injection of mice with

antibody to interferon ox/, (1). Further evidence that

inter-feron was responsible for the antiviral state of peritoneal

macrophagesstemmedfromrecentexperiments in whichan

inverse correlation was observed between the intracellular

levelsof(2' -5')oligo-adenylate(2-5A)synthetaseactivityin

peritoneal macrophages and the permissivity ofthese cells

for VSV (14).Thusthe level of2-5Asynthetase

activity

was

high inresting peritoneal macrophages from adultmice and

low in peritoneal macrophages from mice injected with

antibody to interferon

od/I

(14). These data strongly

sug-gested that injection of mice with antibody to interferon

resultedin theneutralization ofendogenousinterferon

spon-taneously expressed innormal mice. However, most

inves-tigators have not been able to demonstrate biologically

activeinterferonintheperitonealwashingsortissueextr-acts

ofthemajority of mice tested(11,18;J.Galabru,N.Robert,

C.

Buffet-Janvresse,

Y. Riviere, and A. G.

Hovanessian,

J.

Gen. Virol., in press).

The in vivo resistance ofperitoneal macrophages to

sev-eral animal virusespersists when these cellsarefirstplaced

in culture. Although interferon was not recovered in the nutrient medium ofsuspensions offreshly explanted

perito-neal cells, we could demonstrate that these cells conferred an antiviral state on target mouse cell monolayers in coculture experiments. The evidence indicates that

inter-feronwastransferred fromtheperitonealcells to the

under-lying target cells. The results of these experiments are

presented herein.

* Correspondingauthor.

MATERIALS AND METHODS

Mice. Male and female DBA/2, Swiss, C57B1/6, and BALB/c mice 5 to8weeks oldwereobtained from Charles River, Italia S.p.A. (Milan, Italy). Nude mice (BALB/c genetic background) were purchased from Charles River

Breeding Laboratories, Inc. (Wilmington, Mass.) and kept underspecific-pathogen-free conditions during experiments. Preparation of donormousecells. (i) Total peritoneal cells. The peritoneum was washed with 2.5 ml of RPMI 1640 nutrient medium containing 10% fetal calf serum (FCS;

GIBCO Laboratories, Grand -Island, N.Y.).

(ii) Mouse peritoneal macrophages. Mouse peritoneal cells (106 cellsperml)wereallowedtoattachtoapetriplastic dish (Nunc, Roskilde, Denmark) in 10 ml of RPMI nutrient medium containing 10% FCS for 20 min at 37°C. The monolayer was washed three times with RPMI medium without serum to remove nonadherent cells, the remaining cellmonolayerwastrypsinized, thecellswerecounted, and viabilitywas determined. By this technique, almost 50% of the cellsattachedtotheplastic substrate couldberecovered.

(iii) Thymocytes and lymph node cells. Thymocytes and mesenteric lymph node cells were obtained by teasing and filtering the cellsthrough ametallic grid.

Preparation of target cells. For preparation of virus-permissive macrophage monolayers, peritoneal washings were prepared as described above and seeded in 24-well plastic plates (Nunc), each well containing approximately 106 cells perml. Cellswere allowedtoattach totheplastic culture dish at 37°C for 3 h, and nonadherent cells were discarded. The criteria usedto define these firmlyattached cells as macrophages have been previously described (1). Macrophages were cultivated in vitro for 4 to 5 days, at 456

on November 10, 2019 by guest

http://jvi.asm.org/

[image:2.612.57.560.85.239.2]

TABLE 1. Source and activity of the different immunoglobulin preparations

Prepn Source Titer' Source or reference

Anti-mouseIFNb globulin

Sheep no. 1-7(anti-a/I mouseIFN) IRSC,c Villejuif 1.6 x 10-6 (13)

Sheep no. R5/4(anti-ap mouse IFN) Istituto Superiore diSanita,Rome 1.0 X 10-5 E. Proietti, Rome Rabbit anti--y mouse IFN EnzoBiochemicals In., New York 5.0 x 10-3 Catalogno. EAB-473 Ratmonoclonal antibody to mouse fy IFN Trudeau Institute Inc., New York 5.0 x 10-4 (19)

(Havell)

Monoclonal antibody to mousePIFN Erasmus Universiteit, Rotterdam 5.0 x 10-3 (21) (Trapman)

Control hyperimmune globulin

Sheep no. 11 anti-impurities IRSC, Villejuif <1.0 x

1o-1

(13)

Normal serum globulins

Sheep no. 2 IRSC, Villejuif <1.0 x 10-1 (13)

Sheep no. lq Istituto Superiore di SanitA, Rome <1.0 x 10- E. Proietti, Rome

aNeutralizing titer against 4 to 8 U of mouse interferon. bIFN,Interferon.

cIRSC, Institut de Recherches Scientifiques sur le Cancer.

which time they becamepermissive for VSV and EMCV and

were usedastargetcell monolayers. Monolayer cultures of "aged" macrophages showed the same sensitivity as

monolayer cultures of L cellstothe antiviral action ofmouse

interferona113.

MouseL929cells, human Helaand Wishcells, andrabbit RK13 cells were cultivated in RPMI 1640 medium

supple-mentedwith 10% FCS. Chicken embryo fibroblasts (kindly provided by F. Tat6, Rome) were cultivated in Dulbecco

modified Eagle minimum essential medium (Eurobio, Paris) supplemented with 1% chicken serum, 4% FCS, and 10%

tryptose phosphate broth (Difco Laboratories, Detroit, Mich.). Mouse (Swiss) embryo fibroblasts, obtained from 16-day-old Swiss mouse fetuses, were cultivated in RPMI medium supplemented with 10% FCS, 1% nonessential amino acids (GIBCO), and 1% sodiumpyruvate (GIBCO).

Viruses. Theorigin, methodsofpreparation, andassayof VSV (Indiana strain) and EMCV have beendescribed

pre-viously (1).

Hyperimmune and normal serumglobulins. All serawere

decomplemented and extensively absorbedon murinecells. Theimmunoglobulin fractionswereseparated by ammonium

sulfate precipitation (protein concentration varied between 20 and33 mg/ml) and showntobe devoid ofanytoxicity for mouse peritoneal macrophage. The anti-mouse interferon

ax/, globulins didnotneutralize interferon-y.

Thesourcesandactivities of thedifferentimmunoglobulin preparations arepresentedin Table 1.

Preparation of interferon. Mouseinterferon al3 was

pro-duced, partiallypurified (specific activity, 107reference units

permgofprotein), andassayedaspreviouslydescribed(20). One ofthe unitsasexpressedin thetextistheequivalentof four interferon reference units.

Assay for induction of antiviralstatein targetcell

monolay-ers.(i) VSV yieldassay.A0.1-mlsampleof medium contain-ing freshly harvested total peritoneal cells, macrophages, or othercelltypes(usually 2.5 x 106cells)wasfirst mixed with 0.2 ml of soft agar (Bacto-agar [Difco] 0.33% in 5% FCS-RPMI 1640)topreventthedonorperitoneal cells from adhering to the underlying target cell monolayers. After solidification for5minat4°C,asecondlayer of softagarwas applied. After incubation for 24 h at 37°C, agarand donor peritoneal cells were discarded, and the remaining monolayer was washed three times with medium (2%

FCS-RPMI 1640). A 0.2-ml sample of a viral dilution

(mul-tiplicity ofinfection ofapproximately 2) was added toeach

well. After 1 h of incubation at 37°C the cell sheet was

washedthoroughly, and 1ml of nutrientmediumwasadded.

After incubation for18 h at37°Cin a 5%

C02-air

incubator,

thecell-free nutrientmedium wasassayedfor virus in mouse

L929cells.

(ii) Plaque reduction assay. Donor

peritoneal

cells

(usually

5 x 106cells) were mixed in 0.2 ml in soft agar as described

above beforeadding toconfluent cell monolayersin35-mm

tissue cultureplates (Nunc). Asdescribedabove, a second

layerof softagar was

applied,

theculturewasincubatedfor

24 h, the agar containing donor peritoneal cells was

re-moved, thecellmonolayerwaswashed,and agivennumber

ofplaqueforming units was added(2).

RESULTS

Peritoneal macrophages developpermissivity for VSV and EMCV during in vitro culture. VSV and EMCV do not

multiply in the vast

majority

of peritoneal macrophages

freshlyexplanted from4- to7-week-old maleorfemalemice

(1). However,when

peritoneal

macrophageswerecultivated

in vitro for 3 to 5 days, these cells became

permissive

for

VSV(Fig. 1A) or EMCV (datanot shown).

Inanotherexperimentwedetermined the development of

permissivitytoVSV and the levelof2-5Asynthetase

activity

in peritoneal macrophages (DBA/2 mice) duringthe first 5

days of culture. As

previously

noted (14), there was an

excellent inverse correlation between the level of 2-5A

synthetase activity in

peritoneal macrophages

and the

permissivity

of these cells for VSV. Thus, cultivation of

macrophages in vitro for5 days rendered these cells

com-pletelypermissive forVSVandwasassociatedwitha20-fold

decreaseinthelevelof2-5Asynthetase

activity

with respect to

freshly

explanted

peritoneal macrophages (data

not shown).

Freshly harvested peritoneal cellsconfer an antiviral state onunderlying monolayercultures ofagedpermissive macro-phages. Cultivation of macrophages for 5 days rendered these cells

permissive

for VSV (Fig. 1B). The addition of freshly harvested mouseperitonealcells inducedanantiviral state in the underlying macrophage

monolayer

which was clear cut after 7 h ofcocultivation and became more pro-nounced thereafter(Fig. 1B).

on November 10, 2019 by guest

http://jvi.asm.org/

5

4

No

0 3 0

L 2

A

0 1 2 3 4 5 0Z 6b 1U 14 18 22 26 Days ofcultivationinvitro Time (hrs)

FIG. 1. Decayof antiviralstateinperitoneal macrophages during in vitro cultivation (A) and recovery of antiviralstateby cocultivation withfreshly harvested peritoneal cells (B). (A) Peritoneal cells were harvested from theperitoneal cavity of 6-week-old Swiss male mice and seeded in 24-well plastic plates (106 cells per ml ofmedium). Cells wereallowed to fix tothe plastic culture dish at 37°C for 3 h, and nonadherentcells were discarded.Atdifferent days after cellseeding,peritoneal macrophageswereinfected withVSV, and virusyieldswere determined. There werethreemacrophage cultures for each experimental point. (B) Peritoneal cellswereharvested from6-week-old Swiss male mice, andperitoneal macrophageswerecultivated in vitro for 5daystorender thempermissivefor VSV. Peritoneal cells(2.5 x 106,

freshly harvested from theperitoneal cavity of 5-week-old Swissmalemice)wereadded in agarcontaining mediumasdescribed in Materials and Methods for different timesof cocultivation. Agar and donor cells wereremoved, and theunderlying peritoneal macrophageswere infected with VSV. VSVyieldsweredeterminedasdescribed in Materials and Methods. Therewere three macrophage cultures for each point. Symbols: 0,in vitro-aged macrophage monolayer alone; *,peritoneal cells addedtoinvitro-agedmacrophage monolayer.

Table 2 shows theresults of similar

experiments

inwhich

peritoneal cells fromdifferent strains of mice were added to

syngeneic in vitro-aged macrophage cultures permissive for

VSV. In each instance, theaddition offreshperitoneal cells

conferred an antiviral state (to VSV or EMCV) in the

underlying monolayercellculture. Induction ofthisantiviral

state did not occur when antibody to interferon aol was included duringthe periodofcocultivation (Table 2).

TABLE 2. Effect of addition of freshly harvested peritoneal cells fromdifferent strains of mice on the multiplication of VSV or

EMCV in monolayer cultures of agedmacrophages" Virus yield/0.2 ml(logl0meant SE)

Strain Challenge Mouse Mouseperitonealcells,

ofmouse virus Controlb peritoneal anti-Interferona/W

cellsadded added

DBA/2 VSV 3.6

±

0.2

.0.5

3.8

±

0.3

SWISS

VSV

3.3

±

0.3 0.6

±

0.1 3.6

±

0.5

C57B1 VSV 2.5

±

0.2

.0.5

2.7

±

0.2

BALB/c VSV 3.2

±

0.2 1.6

±

0.1 4.1

±

0.2 BALB/c VSV 3.4

±

0.2 1.7

±

0.3 4.2

±

0.4

nude

DBA/2 EMCV 5.3 + 0.3 2.7

±

0.1 5.5

±

0.3

aPeritoneal macrophages explanted from different strains of6-week-old male orfemale mice were cultivated in vitro for 4 days to render them

permissive for VSVorEMCV. Syngeneic fresh peritoneal cells(2.5 x 106) from 5 to6-week-old male mice were added asdescribedinMaterials and

Methods. After 18 h, donor peritoneal cells were removed, and target

macrophage monolayerswereinfected with VSV.Therewere three

macro-phage cultures foreachexperimentalsample.

bTarget macrophage cultures alone.

'Sheep1-7antibodytointerferon (x/ (1:1,000, final dilution).

In the experiments described above, amaximal antiviral

state was conferred whenfreshly harvested peritoneal cells

were cocultivated with permissive macrophage monolayer

cell culturesfor12 to18 h before removal oftheperitoneal

cells in the agar overlay (Fig. 1B). It was of interest to

determinewhetherthisperiodof 12 h was necessary for the

underlyingpermissive macrophage monolayertodevelopan

antiviral state or whether it was necessary for the donor

peritoneal cells tobe present for 12h. Accordingly, mouse

peritoneal cellsin agar wereaddedtoprepared macrophage

cultures,and the donor cellswereremovedatdifferenttimes

thereafter. The underlying macrophage cultures were then

furtherincubatedat37°Cfor 12 hbefore infectionwith VSV.

Theresults ofthese

experiments

showedclearlythata5-min

period of cocultivation offreshly

peritoneal

cells with

pre-pared macrophage cellmonolayerswassufficienttoinducea

2-log1o

reduction of VSV yield (compared with control

macrophages alone).Thisantiviralstate didnot occurwhen

antibody to interferon

at/l

was added to the cell mixture

(data notshown).

Freshly harvested mouseperitoneal cells conferanantiviral state on target mouse cells but not on heterologous cells. Freshlyharvested mouse peritoneal cellsconferred an

anti-viral state on target monolayer cultures of in vitro-aged

mouse macrophages,mouseembryo fibroblasts, andmouse Lcells(Table3). This effect was abrogated bythepresence

ofantibody to mouse interferon

a/c

during the period of

cocultivation. Mouse peritoneal cells did not confer an

antiviral state on targetmonolayerculturesofheterologous

chicken embryo fibroblasts, rabbit kidney cells, or human Wish cells. A minimal effect may have occurred in HeLa cells.

on November 10, 2019 by guest

http://jvi.asm.org/

[image:3.612.141.473.71.293.2]

[image:4.612.53.557.82.308.2]

TABLE 3. Freshly harvested mouse peritoneal cells confer an antiviral state on target mouse cells but not on heterologous cells

VSV yield/0.2 ml MeanVSV plaques/

Addition Cellmonolayertarget (log1omeant SE)' dish(%of

control)b

None Aged mouse macrophage monolayers (5 3.3 ± 0.2

days)

Peritoneal cells 1.0 ± 0.1

Peritoneal cells, antibodyc 3.4 ± 0.3

None Mouse embryo fibroblasts 5.5 ± 0.2 21.5

Peritoneal cells 2.5 + 0.1 2.5 (11)

Peritoneal cells,antibody 6.0 ± 0.4 17.0 (79)

None MouseLcells 52.0

Peritoneal cells 11 (21)

Peritoneal cells, antibody 53 (102)

None Chicken embryo fibroblasts 5.9 ± 0.3 15.5

Peritoneal cells 5.8 ± 0.2 13.5 (87)

Peritoneal cells, antibody 6.0± 0.2 13.5(87)

None Rabbit kidney (RK13) 5.5 ± 0.1 32.5

Peritoneal cells 5.7 ± 0.3 32.5(100)

Peritoneal cells, antibody 5.7 ± 0.1 36.5(112)

None HeLa cells 5.7 ± 0.3 9.0

Peritoneal cells 4.5 + 0.4 5.0(56)

Peritoneal cells, antibody 5.7 ± 0.1 8.5(94)

None HumanWish cells 52.0

Peritoneal cells 38.5(74)

Peritoneal cells, antibody 45.0(87)

a Different types of cellmonolayers (permissive for VSV) were cocultivated with 2 x 106peritoneal cells(freshly harvested from 5-week-oldSwiss female mice). After 18 h at 37°C the donor cells in agar were removed, and the target cell monolayers were infected wih VSV. There were three infectedculturesfor each experimental sample.

bDifferenttypes ofcell monolayers were cocultivated with 6x106peritoneal cells(freshly explanted from 5-week-old Swissmalemice). After18 h at37°C cell

monolayerswerewashed and further processed forVSV plaque assay as described in Materials and Methods. There were twocellcultures for eachexperimental sample.

Sheep 1-7 antibody to interferonal/P(1:500,final dilution).

Relationship between the number ofmouseperitoneal cells

embryo

fibroblasts

(Fig. 2A).

An antiviral state did not

and theircapacity to confer an antiviralstate.Amounts of 2 x developinmousefibroblastswhen

antibody

tointerferona/I

106to 3 x 106peritonealcells wererequiredtoinducea50% wasincorporated into thenutrientmedium

regardless

of the

inhibition ofVSV plaques in monolayer cultures ofmouse numberof

peritoneal

cells added.

N.ofperitoneal cells per dish(log,O) N.4.5ofperitoneal5 macrophages5.5 perwel6 (log,,)6.5

FIG. 2. Relationship between the number of mouse peritoneal cells and their capacity to confer an antiviral state. (A) Various numbers oftotal peritoneal cells (harvested from 5-week-old Swiss female mice) were cocultivated with confluent monolayers ofmousefibroblasts. After 18 h at 37°C, peritoneal cells were removed, and the fibroblast monolayers were processed for VSV plaque assay. The values are

expressed as the percentage of plaques (mean of four individual cultures) with respect to the mean numberof plaquesincontrol culturesof untreated fibroblasts (58 + 4.1). Symbols: 0, in presence of antibody to mouse interferona/l during the cocultivation period(sheep 1-7; 1:1,000, final dilution). 0, in absence ofantibody during the cocultivation period. (B) Peritoneal macrophages, freshly harvested fromn 5-week-old Swiss female mice, were separated from other peritoneal cells by adherence to the plastic substrate. Aftertrypsinization, different numbers of these macrophages were suspended in agar containing medium and cocultivated with 4-day-old monolayer cultures of macrophages. After 18 h at 37°C the donor cells in agar were removed, and the macrophage monolayers were washed and infected withVSV.

There were three infected cultures for each point.

on November 10, 2019 by guest

http://jvi.asm.org/

[image:4.612.136.471.424.631.2]

[image:5.612.63.561.84.279.2]

TABLE 4. Capacity ofdifferent celltypes to conferanantiviral state ontarget mousecell monolayer cultures

MeanVSVplaques! VSV

yieldl0.2

ml Expt Donorcells Targetcells dish(%ofcontrol) (Logl0 mean

± SE)

1" None Mousefibroblasts 76(100)

Mouse peritonealcells Mousefibroblasts 0

Mouse peritoneal cells, antibody to IFNb Mousefibroblasts 78(103)

Mouse thymocytes Mouse fibroblasts 33(43)

Mouse thymocytes, antibody to IFN Mouse fibroblasts 90(118)

2C None Aged in vitro mouse macrophage cultures 2.8± 0.3

Mouse peritoneal cells Agedin vitromousemacrophage cultures c0.5

Mouseperitoneal cells,antibodyto IFN Aged in vitro mouse macrophage cultures 2.6± 0.1 Guineapig peritoneal cells Agedinvitromousemacrophage cultures 2.5± 0.1 Guineapig peritoneal cells, antibody toIFN Aged in vitro mouse macrophage cultures 2.5± 0.3 Rabbit peritoneal cells Aged in vitromouse macrophage cultures 3.0 ± 0.2

3d

None Agedin vitromouse macrophage cultures 4.5 ± 0.1

Mouseperitoneal cells from normal Aged invitromouse macrophage cultures 1.8 ± 0.4 untreated mice

Mouse peritonealcellsfrommice Aged in vitro mouse macrophagecultures 4.9 ± 0.4 preinjectedwith antibodyto IFN

Peritonealcells andthymocyteswereexplanted from 6-week-old Swissmale mice, and 5 x 106 cellswerecocultivatedwith mouseembryo fibroblastsas

describedinMaterials and Methods. Insome cell cultures antibodyto mouseinterferon ox/p(sheep 1-7;final dilution, 1:500)was added.After 18 h at37°C fibroblast monolayerswerewashedandfurtherprocessed forVSVplaque assay.There werefour dishes foreachexperimental condition.

bIFN, Interferon.

' Peritoneal macrophages (from 6-week-old Swissmalemice)cultivated in vitro for4daysweretreated with2 x 106 peritoneal cells as indicated. After 18 h at

37'Cperitoneal cellswereremoved,and cellmonolayerswereinfectedwithVSV. Therewerethreeinfectedcell cultures for eachgroup.

"Peritoneal macrophages (from 7-week-old Swiss male mice) cultivated in vitro for4daysweretreatedwith 2.5 x 106 peritoneal cellsfreshlyharvestedform 8-week-oldSwissmale mice(either untreatedorpreinjectedintraperitoneallywithantibodytointerferon4daysbefore [sheep1-7;diluted1:10,0.2 ml permouse]). After18 h at37°C, peritoneal cellswereremoved, and cell monolayers wereinfectedwithVSV.There were three infected cellculturesfor eachgroup.

Figure 2Billustrates the results of an experiment in which

weused different numbers of freshly harvested mouse peri-toneal macrophages (separated from those periperi-toneal cells which did not adhere firmly to the plastic substrate). Over

98% of the attached cells were stained for nonspecific

esterase(1). When these cells were detached by trypsin and cocultivated with aged monolayer cultures of macrophages

permissivefor virus replication, less than105donor cells was

sufficient to induce 50% inhibition of VSV yield. These

results indicated that peritoneal macrophages constitute a veryefficient cell population for the transfer of the antiviral state.

Type of cells capable of conferring an antiviral state on mousecellmonolayers permissive forVSV. It wasof interest todetermine whether other types offreshlyexplanted mouse

cells and cells from other species could also confer an

antiviralstate on mousecell monolayers. Mouse thymocytes

conferredonly aslightinhibition on mouse fibroblasts

com-paredwith mouseperitoneal cells (experiment 1 in Table 4).

Likewise, mouse mesenteric lymph node cells conferred

only a slight antiviral state on mouse fibroblasts (data not

shown). In contrast, heterologous guinea pig and rabbit

peritoneal cells (experiment 2 in Table 4) did not induce an

antiviral state in underlying mouse macrophage monolayer cultures. Likewise, cocultivation of RK13 cells with mouse embryonicfibroblasts was notassociated with the induction

ofan antiviral state (data not shown).

Peritoneal cells from mice preinjected with antibody to

interferon did not confer an antiviral state on underlying target mousemacrophage monolayer cultures (experiment 3 in Table 4).

Induction of the antiviral state by mouse peritoneal cells is specifically abolished by antibodies to mouse interferon

a/P.

The above experiments showed that the capacity of perito-neal cells to confer an antiviral state on target monolayer

cultures of mouse embryonic fibroblasts or aged mouse macrophages was abrogated when sheep 1-7 antibody to interferon

o/,I

wasincorporated in the coculture medium. It was ofinterest to test the specificity of this effect of antibody and to determine the minimal amount of neutralizing anti-body required. Two different sheep antibodies to mouse interferon a/1 (including a purified fraction) were highly effective in abrogating the capacity of mouse peritoneal cells to induce the antiviral state in target mouse cells, whereas polyclonal rabbit and monoclonal rat antibody to mouse interferon y, sheep antibody to the impurities in the inter-feron

oJ/I

preparations, and normal sheep immunoglobulin were ineffective (Table 5). A rat monoclonal antibody to mouse interferon fi exerted some neutralizing effect, al-though this effect was much less marked than that afforded by the same number of neutralizing units of antibody to interferon ox/,.

Figure 3 shows the relationship between the amount of neutralizing antibody of sheep 1-7 in the coculture and the extentof the antiviral state induced by donor peritoneal cells

onmonolayercultures ofmouseembryonic fibroblasts(Fig.

3A) and macrophages (Fig. 3B). DISCUSSION

Resting peritoneal macrophages from normal mice were resistant to VSV and EMCV in vivo (1) and when first placed in culture, as shown herein. The evidence indicating that interferon was responsible for this antiviral state may be summarizedasfollows. First, peritoneal macrophages taken from mice injected with antibody to mouse interferon ot/l

(but not injected with control immunoglobulins) were fully permissive for these viruses (1). Second, 2-5A synthetase activity was considered a sensitive marker of interferon activity (14), and we found an excellentinverse correlation between the level of2-5A synthetase activity of peritoneal

on November 10, 2019 by guest

http://jvi.asm.org/

[image:6.612.56.299.105.276.2]

TABLE 5. Effect of different antibodies on the capacity of peritoneal cells to induce an antiviral state in peritoneal

macrophages permissive forVSVa

Donor VSVyield/0.2 ml(logl,

peritoneal Antibody presentin cocultures meant SE) cells

Expt1" Expt 2' 2.5 ± 0.2 2.5 ± 0.2

+ '0.5 '0.5

+ Sheep1-7antibody to

IFNd

do/, 2.6 ± 0.4 2.6 ± 0.1 + Sheep1-7purified fraction 3B 3.5 ± 0.5

(14)

+ SheepR 5/4 antibodytoIFNa/ 3.1 + 0.5

+ Ratmonoclonalantibody to 1.5 ± 0.5 IFN P

+ Rabbitantibody to IFN y -0.5

+ Ratmonoclonal antibody to 0.6 ± 0.1 IFN y

+ Sheep11antibodytoimpurities .0.5 ininterferon preparations

+ Normal sheepimmunoglobulin 0.5

'Peritoneal macrophages explanted from5 to 6-week-old Swissmalemice

werecultivated invitro torender them permissiveforvirus replicationand usedastargetmonolayer cultures. After4days.macrophageswere

coculti-vatedat37°Cwith2 x106peritoneal cells freshly explantedfrom5-week-old Swiss male mice. The neutralizing titers of the anti-interferon globulin preparations(finaldilutionintheexperimental test)wereasfollows.

Experi-ment1:anti-a/13 mouseinterferon globulin (sheep1-7),2 x 103against 8 Uof

a/1 mouseinterferon; anti-ax/1 mouseinterferonglobulin (sheepR5/4). 1 x

103 against 8 U ofcx/1 mouse interferon; rabbit anti-mouse y interferon globulin, 1x 103against 8U of mouseyinterferon. Experiment2:anti-a/,8

mouse interferon globulin (sheep 1-7), 4 x 10' against 8 Uofc/P13 mouse

interferon;anti-a/13mouseinterferon globulin (sheep1-7;fraction3B), 3x103

against 8 U of cs/1 mouse interferon; antibody to y mouse interferon (monoclonalantibody),2.5x 103against10Uof mouseyinterferon;antibody

to1 mouse interferon (monoclonal antibody), 5x 102against 5U of1Bmouse

interferon. After18 hofcocultivation at37°C,peritonealcellswereremoved,

andmacrophageswereinfectedwithVSV. Virusyieldsweredeterminedas

described in MaterialsandMethods.

hThere were three infected macrophage cultures for each experimental

condition.

'There were five infected macrophage cultures for each experimental condition.

dIFN,Interferon.

macrophages and their permissivity for VSV (see above).

Nevertheless, we have been unable to recover biologically

active interferon in theperitoneum ofnormal miceorinthe

nutrient medium of cultures of peritoneal macrophages.

Galabru and co-workers (in press) reported low levels of

circulatinginterferon in14of 160 normalmice.The

question

remains, therefore, whether interferon is present in most

normalmice and whether it isresponsible fortheseemingly

inherent antiviral state ofresting peritoneal macrophages

andprobablyother cells ofthenormalmouse.

Thefollowingresults presented hereinarguestronglythat

interferonisresponsible for this antiviralstate.

(i)Mouseperitoneal cellsareresistanttoVSV andEMCV

when first placed in culture and show high levels of2-5A

synthetase activity. After 4 to 5 days in culture these cells become permissive to these viruses, and the level of 2-SA

synthetase activity

declines.

(ii) Freshly harvested mouseperitoneal cells (and proba-bly macrophages constitute the most effective population;

Fig. 2B)canefficientlytransfer this antiviralstate to

permis-sive target cells, provided they are also of mouse origin

(monolayer cultures of aged macrophages, embryonic

fibro-blasts, or L cells). The antiviral state cannot be conferred

efficiently or at all on heterologous cells known to be

insensitive to mouse interferon, such as chicken embryo,

and rabbit kidney, and human HeLa and Wish cells (Table

3).

(iii)Thekinetics of transfer of the antiviral state(Fig. 1B)

are in accord with previous observations by Dianzani and Baron(8) on thetime necessary for exogenous interferonto induce a full antiviral state in susceptible cells.Furthermore,

the experiments show that cocultivation ofperitoneal cells

with target cellsforonly a few minutes suffices to transferan antiviral state, provided that sufficient timeis subsequently

permitted forthe development of the antiviral state in the

target cells.

(iv) Transfer of the antiviral state was in all instances

completely abolished by the addition ofantibody to

inter-feron al/1 in the mixed cell cultures.

(v) Heterologous guinea pig or rabbit peritoneal macro-phages cannot transfer an antiviral state to mouse cells (Table 4).

(vi) Peritoneal cells from mice injected with antibody to interferon ao/1, which are permissive for VSV (1), cannot transfer the antiviral state to target mouse cells (Table 4).

The failure of two different preparations of antibody againist mouse interferon y to inhibit the phenomenon, the incomplete effect of antibody to interferon 1, and the com-pleteinhibition afforded by two preparations of antibody to interferon ac/ indicate that the transfer of the antiviral state is mediated by interferon a or

P

or both, but not by interferon

y.

This ensemble of results suggests that mouse peritoneal cells contain biologically active interferon (either intra-cellularly oratthe surface) and that these cells transfer this interferon to susceptible mouse cells. Furthermore, the experimental resultsindicate clearly that the peritoneal cells transfer interferon rather than inducing the production of interferon in the target cells. Nevertheless, it is somewhat surprising to us that, in view of the extent of the antiviral effect induced in the target cells by peritoneal cells resus-pended in agar, we have not been able to recover any interferon in the nutrient medium of cocultures. It seems possible thatthe mechanism of transferof the antiviral state by macrophages is more effective than might be suspected, as was suggested many years ago by Glasgow (9).

There have been a few reports of the transfer of the antiviral state in coculture. In the experiments of Blalock and co-workers (3-5), the transfer of the antiviral state occurred between interferon-treated cells and heterologous targetcells and was notinhibited by the presence of antibody to interferon. Their phenomenon appears to be totally dif-ferent, therefore, from the one we have described. Ito and co-workers (15) showed that mouse peritoneal cells from normal orgermfree mice alsotransferred an antiviralstateto homologous mouse cells (but not to heterologous human cells), and the establishment of this antiviral state was partially blocked by the presence of antibody to type I interferon. They concluded, therefore, that interferon was the responsible factor. Soldateschi and co-workers (D. Soldateschi, D. Boraschi, and A. Tagliabue, Antiviral Res., in press) found that co-culture of mouse peritoneal cells and mouse embryonic fibroblasts resulted in an inhibition of EMCV multiplication, but they concluded that this effect was not mediated by interferon because there was not a direct correlation between anti-EMCV activity and inter-feron production. The recent experimental results of Vogel and Fertsch (22) are inaccordwith the observation ofIto et al. (15) and ourselves. They showed that the treatment of C3H/HeN mouse peritoneal macrophages in vitro with

on November 10, 2019 by guest

http://jvi.asm.org/

A

.n

>n 1

B

Neutralizingtiterofantibody(logO) titerofantibody (loglo)

FIG. 3. Capacity of different amounts of antibody to interferon ot/, to abrogate the antiviral state in mouse target cells induced by peritoneal cells. (A) Confluent monolayers of mouse embryo fibroblasts were cocultivatedwith 5x 106 peritoneal cells (freshly explanted from 5-week-old Swiss male mice) in the presence of different amounts of antibody to interferon a/,B. After 18 h at 37°C, peritoneal cells were removed, and mouse fibroblasts were processed for VSV plaque assay. There were four fibroblast cultures for each experimental condition. Thevalues are expressed as percentage of plaque (means of four individual cultures) with respecttothe meannumber of plaques in control untreatedfibroblasts (42±2.5). No plaqueswere detected in fibroblast cultures cocultivated withperitonealcells in theabsence of antibody. Theratio of peritoneal cells to embryonic fibroblasts was2.5:1.(B)Peritoneal macrophages explanted from 8-week-old Swiss male mice were cultivated in vitrotorender thempermissive for VSV. After4days, macrophages were cocultivated with 3 x 106 peritoneal cells, freshly explanted from 6-week-old Swiss male mice,inthe presence ofdifferent amounts of antibodytointerferon

a/P.

After 18hat37°C,peritoneal cells were removed, and macrophage cultures were infected with VSV. There were three macrophage cultures for each experimental condition. The ratio of peritoneal cellstomacrophage cultureswas 10:1.

bodytointerferonot/agreatly reduced the capacity of these

macrophages to phagocytose opsonized erythrocytes.

Fur-thermore, they demonstrated that this phagocytic activity

could be transferred in coculture from C3H/HeN

macro-phages to macrophages from endotoxin-hyporesponsive

C3H/HeJ miceandconcluded thatthiseffect wasmediated

by interferon (22).

Although thereis accumulating evidencethatinterferon is spontaneously expressed under some conditions in normal

hunians(16), rabbits(7),and mice(1,11, 14, 15, 22; Galabru

et al., in press), several important questions remain to be

answered. Asregards mice, dothe peritoneal cellsproduce

interferon or do they bind interferon which is produced

elsewhere?Isinterferonproducedconstitutively byavariety

of host cells in normal mice, or is it continually induced

because ofthebacterial florainmicein thealimentarycanal

assuggestedby theexperiments ofGalabruetal. (inpress)?

Whatever the mechanism, the presence of low levels of

interferon in normal animals appears to be an integral componentin the hostdefenseagainstviralinfection. It may well be that these low levels of interferon before viral infection arefar moreimportant for the host than the much greater amounts of interferon which are produced in the course of viral infection. Likewise, it is possible that low levelsofendogenous interferon alsoexertothereffects in the

animal, such as restricting the growth of neoplastic cells.

Thus, injection of mice with antibody to interferon

ct/n

markedly enhanced the tumorigenicity of xenogeneic (17)

andsyngeneic(11) tumor cells. It remains to be determined

whether these low levels of interferon also play a role in regulating the function of certain host cells, as well as the

expression of certain surface antigens,suchas

histocornpat-ibility antigens, which are known to be affected by

exogenousinterferon(10). Ourexperimental resultssupport

a hypothesis proposed by Bocci (6) of a "physiological

interferon response" whichismanifested bytheproduction

of interferon "at specific sites" and which "can maintain active defense systems essential for survival."

ACKNOWLEDGMENTS

This work was supported in part by grants from Consiglio Nazionale delle Ricerche (Progetto Finalizzato Controllo delle Malattie da Infezione No. 83.02916.52 and Progetto Finalizzato OncologiaNo. 84.00498), theRichardLounsbery Foundation, the SimoneetCinodel DucaFoundation,theDirection de Recherches et Techniques (contract 82-420), the Association pour le Develop-pement de la Recherche sur le Cancer at Villejuif, the Institut National delaSanteetdela RechercheMedicale (contract 127 012), and the Foundationpourla RechercheMedicale

Frangaise.

We are indebted to the skillful technical assistance of L. Santurbano.

LITERATtJRECITED

1. Belardelli, F., F. Vignaux, E. Proietti, and I. Gresser. 1984. Injectionofmicewithantibodytointerferonrendersperitoneal macrophages permissivefor vesicularstomatitis virus and en-cephalomyocarditis virus. Proc. Natl. Acad. Sci. USA 81:602-606.

2. Benedetto, A.,G. B.Rossi,C. Amici,F. Belardelli,L.Cioe,G. Carruba, and L. Carrasco. 1980. Inhibition of animal virus products by means of translation inhibitors impermeable to 2_c

0 u

z

NZ

on November 10, 2019 by guest

http://jvi.asm.org/

[image:7.612.163.456.70.295.2]

normal cells. Virology 106:123-132.

3. Blalock, J. E., and S. Baron. 1977. Interferon-inducedtransfer of viral resistance between animal cells. Nature (London) 269:422-425.

4. Blalock, J. E., and S. Baron. 1979. Mechanisms of interferon-induced transfer of viral resistance between animal cells. J. Gen. Virol. 42:362-372.

5. Blalock, J. E., and G. J. Stanton. 1978. Efficient transfer of interferon-induced virus resistance between human cells. J. Gen. Virol. 41:325-331.

6. Bocci, V. 1981. Production and roleof interferon in physiologi-calconditions. Biol. Rev.56:49-85.

7. Bocci, V., M. Muscettola, L. Paulesu, and G. Grasso. 1984.The physiological interferon response. II. Interferon is present in lymph but not in plasma of healthy rabbits. J. Gen. Virol. 65:101-108.

8. Dianzani, F., and S. Baron. 1975. Unexpectedlyrapid action of humaninterferon inphysiological conditions.Nature(London) 257:682-683.

9. Glasgow, L. A. 1970. Transfer ofinterferon-producing macro-phages: new approach to viral chemotherapy. Science 170:854-856.

10. Gresser, I. 1984. The effect of interferonontheexpression of surface antigens, p. 113-132. In J. Vilcek and E. De Maeyer (ed.), Interferons and the immune system. Elsevier/North-HollandPublishing Co., Amsterdam.

11. Gresser, I., F. Belardelli, C. Maury, M. T. Maunuory, and M.G. Tovey. 1983.Injection of mice with antibodytointerferon enhancesthegrowth oftransplantable murine tumors.J. Exp. Med. 158:2095-2107.

12. Gresser, I., C. Bourali, M. T. Thomas, and E. Falcoff. 1968. Effect of repeated inoculation of interferon preparations on infection of mice with encephalomyocarditis virus (32723). Proc. Soc. Exp. Biol. Med. 127:491-496.

13. Gresser, I., M. G. Tovey, M. T. Bandu, C. Maury, and D.

Brouty-Boye. 1976. Role of interferon in the pathogenesis of virus diseases in mice as demonstrated by the use of anti-interferon serum. I. Rapid evolution ofencephalomyocarditis virusinfection. J. Exp. Med. 144:1305-1315.

14. Gresser, I., F. Vignaux, F. Belardelli, M. G. Tovey, and M. T. Maunoury. 1985. Injection of mice with antibody to mouse interferon a/,Bdecreases the level of 2'-5'oligo-adenylate syn-thetase inperitonealmacrophages.J. Virol.53:221-227. 15. Ito, Y., H.Aoki, Y. Kimura, M. Takano, K. Shimokata, and K.

Maeno.1981. Naturalinterferon-producingcells in mice. Infect. Immun.31:519-523.

16. Lebon, P., S. Girard, F. Thepot, and C. Chany. 1981.Presence constante de interferon a dans les liquides amniotiques humains.C.R. Acad. Sci.293:69-71.

17. Reid, M. L,N. Minato,I. Gresser, J. Holland, A. Kadish, and B. R.Bloom. 1981. Influence of anti-mouse interferon serum on thegrowthandmetastasis oftumorcells.Proc.Natl.Acad. Sci. USA78:1171-1175.

18. Sen, G. S. 1982. Mechanism of interferon action: progress toward itsunderstanding. Prog. Nucleic AcidRes. Mol. Biol. 27:105-156.

19. Spitalny,G. L., and E. A. Havell.1984. Monoclonal antibodyto murinegammainterferon inhibitslymphokine-induced antiviral and macrophage tumoricidal activities. J. Exp. Med. 159:1560-1565.

20. Tovey, M.G., J. Begon-Lours, and I. Gresser. 1974. Amethod forthelarge scaleproduction ofpotent interferonpreparation. Proc. Soc. Exp. Biol. Med. 146:809-815.

21. Trapman, J. 1983. Distinctantigenic characteroftwo compo-nentsofpoly(I)-poly(C)-inducedmouse Lcellinterferon. FEBS Lett. 109:137-140.

22. Vogel, S. N., and D. Fertsch. 1984. Endogenous interferon production by endototoxin-responsive macrophages provides an autostimulatory differentiation signal. Infect. Immun. 45:417-423.

on November 10, 2019 by guest

http://jvi.asm.org/