Human immunodeficiency virus infection of monoblastoid cells: cellular differentiation determines the pattern of virus replication.

Human

Immunodeficiency Virus

Infection of Monoblastoid

Cells:

Cellular Differentiation

Determines

the Pattern

of Virus

Replication

C. DAVID

PAUZA,1*

JOSEGALINDO,'AND DOUGLAS D. RICHMAN2

Developmental Biology Laboratory, the Salk Institutefor Biological Studies, P.O. Box85800, SanDiego,

California

92138-9216,1

andDepartments of PathologyandMedicine, University ofCalifornia,

and Veterans Administration MedicalCenter, San Diego, California 921612

Received 21 March 1988/Accepted 27 June 1988

Stringent control of human immunodeficiency virus (HIV) replication was observed in the human

monoblastoid cell line U937. A low-multiplicity infection of these cells by the LAV1 strain of HIV was

productive for 2.5 days; then virus replicationbecame restricted andnofurther evidence of virionproduction

was observed. The dramatic decrease in HIV production was due in part of reduced accumulation of

cytoplasmic viralRNA and occurred in the absence of evident cytopathic effects. In contrast, infected cells

inducedtodifferentiate by phorbolester, vitaminD3, orlymphokinesupernatant did not releasemarkers of

HIVdespite the accumulationofsignificant levelsofcytoplasmicviral RNA.HIVinfectionalteredthepattern

ofc-mycRNA accumulation in U937 cells.Expressionof thisgenechanges normallyinresponseto the stateof cellulardifferentiation; ininfected cells the level ofc-mycexpression wascorrelatedto the levels of viralRNA

accumulation and not to cellular differentiation. These results suggest that restricted replication of HIV in

monocytesmightbeanimportantmechanism of virus persistence and demonstratearelationshipbetweenHIV

replication and monocyte differentiation.

Accumulating evidence indicates that infection of

mo-nocytes by human immunodeficiency virus (HIV) is an

importantcomponentinthepathogenesisofacquired

immu-nodeficiency syndrome (AIDS). HIV-infected individuals

manifest elevated levels of circulating lysozyme (19), which

is released by activatedmonocytes (8, 18); increasedserum

lysozyme levels are correlated with progression to AIDS

(19). Monocytes derived from AIDS patients have

signifi-cantly reduced phagocytic (9) and chemotactic (35, 40)

activities, and thepercentageofthese cells thatareHLA-DR

positive is approximately halfthe level observed normally

(42). The monocytephenotypes observed in AIDS patients

suggest thatactivation and differentiation of these cells are

altered byHIVinfection.

In addition to the phenotypic alterations observed in

monocytepopulations derived from AIDS patients, evidence

exists that these and related cells aredirectly infected with

HIV. Viral RNA and antigens have been observed in brain

tissue macrophages (14, 26, 43) and in the macrophagelike follicular dendritic cells inlymph nodes (2, 3, 31). Cellswith

macrophage markers inthe skinand lungs of AIDS patients

have been shownto contain infectious virus, viral antigens,

or viral nucleic acids (6, 23, 31, 34). Peripheral blood

monocytes arepermissive forHIVin vivo(15, 23, 30), and

virus can be isolated from these same cell populations in

infected individuals (15, 23). HIV infection ofmonocytes,

and the variety of specialized cell types related to this

lineage, constitutes a significant component of virus

distri-bution inthe infected individual.

The role of monocytes in AIDS is controversial. It has

been suggested that these cells serve as the reservoir for

virus in the body (22, 25). The observation oflow ratesof

HIV replication in monocytes in vitro and the absence of

cytopathic effects on these cells has beenproposed as one

possible explanation for HIV persistence in vivo(15, 21). It

hasalso been proposed(33) thatinfectedmonocytesserveas

*Corresponding author.

theprincipalagents of HIV disseminationto helperTcells,

because thesetwo cell types necessarilycome into contact

during the course of antigen presentation and the T cell

receives activation signalsinconsequenceofthis interaction

(33). New insights into the role of monocytes in AIDS are

especially important in light of the recent discovery that

dideoxynucleosides, which constitute animportant class of

agents for the treatment ofAIDS, have reduced antiviral

activityin cells of thislineage (39).

The interaction between HIV and the host monocyte is

thoughttobecomplex, involving specialmechanisms for the

control of virus production. Accordingly, we sought to characterize thebiologyof HIVreplicationinmonocytes. In these initial studies,the humanmonoblastoid cell line U937 (41)wasusedas amodel forHIV-monocyteinteractions and

the relationship between virus production and cellular

dif-ferentiation. Weobservedthat HIVreplicationin these cells

was regulated stringently. A low-multiplicity infection of

U937 cellswas productive initially. At 2days postinfection,

we observed that cytoplasmic viral RNA, cell-free reverse

transcriptase activity,andp24gagantigenlevels diminished

rapidly withoutevidentcytopathic effects. Thus, productive

infection had become restrictedand virus release ceased. In contrast, infected U937 cells induced to differentiate by treatmentwithaphorbolester(tetradecanoylphorbolacetate

[TPA]), 1,25 dihydroxyvitaminD3,orlymphokine

superna-tants did not release markers of HIV despite significant

accumulation ofviral RNA inthe cytoplasm of TPA-treated

cultures.

Direct effectsof viralinfection on the cellular phenotype

were assessed by examining the cellular gene c-mycthatis

expressed normallyinU937 cells. Thepatternofc-mycRNA

accumulation was altered in infected cells as opposed to

uninfectedcontrols, attestingtoadirect effect ofHIVonthe

phenotype ofthese cells. Thealtered patternofc-mycRNA

accumulation in infected cells indicates that these changes

are mediated at least in part by direct effects of HIV on

cellulargene expression.

3558

0022-538X188/103558-07$02.00/0

Copyright C 1988,American Society forMicrobiology

on November 10, 2019 by guest

http://jvi.asm.org/

MATERIALS AND METHODS

Cellculture andpreparation of virusstocks. U937 cells (41)

weremaintained at densities of between 1 x

105

and 5 x 105

cells per ml in RPMI 1640 medium with 10% heat-inactivated

fetal bovine serum, supplemented with

penicillin-streptomy-cin-glutamine. The cells were passaged on the day of before

infection. Cell-free stocks of the

LAV1bl,

strain of HIV (1)

wereprepared in the T-lymphoblastoid cell line, CEM. The

titers were adjusted to 107 tissue culture infective doses

(TCID50s)

per ml, asdetermined by syncytiumformation in

aterminal dilution assaywith MT-2 cells(20).

Virus production in differentiated and undifferentiated

U937cells. Cultures of U937 cells were infected at a

multi-plicity of 0.2 TCID50 per cell. At 8 h later, the culture was

split and TPA was addedto half, at a final concentration of

1 ,ug/ml. After 1 h, cells were recovered from both cultures

bycentrifugation, washed once inphosphate-bufferedsaline

(PBS), and suspended in fresh medium to a final

concentra-tion of 5x 105cells perml. The flasks, each containing 20 ml

of culture, were then incubated at 37°C in an humidified

incubator with 5% CO2 in air and were harvested at 0, 1, 2.5,

and 5 days after TPA treatment. The cultures were

supple-mented with an equal volume of fresh medium at 2.5 and 4

dayspostinfection. The effects of 1,25 dihydroxy vitamin D3

or lymphokine supernatant treatments were assessed in a

similar manner; vitamin D3 was added to infected cell

cultures at a final concentration of

10-9

M (provided by

Milan Uskokovick, Hoffmann-La Roche Inc.), and

commer-cially available delectinated supematant form

phytohemag-glutinin-activated human peripheral blood lymphocytes

(Ad-vanced Biotechnologies) was added as a 20% volume

supplement to infected U937 cells in culture.

At each time point, the cells were collected by

centrifu-gation. A 1-ml sample of the cell-free supernatant was

retained and stored at -70°C for assay of reverse

transcrip-tase activity and quantitation of p24 gag antigen (p24

en-zyme-linked immunosorbent assay, Abbott Laboratories,

Chicago, Ill.). The flasks were rinsed with 5 ml of PBS

containing 0.5 mM EDTA to recover adherent cells. The

harvested cells were washed again with cold PBS and

suspended in 2.5 ml of 15 mMNaCl-15 mM Tris

hydrochlo-ride (pH 7.5). Then, 100,ul of 20 mM vanadyl ribonucleoside

and 200 ,lI of 20% Nonidet P-40 were added. Nuclei were

removed by a2.5-min spin at 2,500 rpm in a Beckman TJ-6

centrifuge, and the supernatant was combinedwith anequal

volume of 150 mM NaCl-150 mM Tris hydrochloride (pH

7.5)-i

mM EDTA. Proteinase K (Boehringer Mannheim)

was then added to a final concentration of 200 ,ug/ml.

Samples were incubated for 45 min at 56°C and then

ex-tracted with an equal volume of 1:1 phenol:chloroform

containing 0.1%hydroxyquinoline-0.5% 2-mercaptoethanol.

The aqueous phases were recovered, and RNA was

precip-itated by the addition of sodium acetate (pH 5) to a final

concentration of 0.3 M plus 2.5 volumes of ethanol (10).

Samples were stored at -20°C

overnight,

and the nucleic

acid precipitates were recovered by centrifugation.

The RNA pellet was purified further to remove

contami-nating vanadyl ribonucleoside (10). The dried

pellet

was

suspended in 250 ,u of3 M sodium acetate

(pH

5), and the

RNAwas recoveredby brief

centrifugation

inan

Eppendorf

microcentrifuge. The pellet was rinsed in 70%

ethanol,

suspended in water, and precipitated again. The

resulting

material was of sufficient purity that the concentration of

RNA could be determined from the optical

density

at 260

nm. Samples of eachpreparation werealso runon agarose

gelsas an additional check on RNA contentand integrity.

The reversetranscriptaseactivity of cell-freesupernatants

wasdetermined by amodification-ofthe methodofPopovic

etal. (36). Aftera2-hincubation at37°Cin thepresenceof

5 mM

MgCl2-poly(rA)

template-oligo(dT) primer, with

[32P]dTTP as label, portions ofthe reaction mixture were

spottedonDE81 paper andallowedtodry

briefly.

The paper

was washedtwiceat roomtemperature in 2x SSC (1x SSC

is0.15 MNaCland 0.015 M sodium citrate), rinsedoncein

95% ethanol, andthenexposedtofilm for3 to 5 h. Individual

sections werethencutout,andthe

incorporated

radioactiv-itywasdeterminedby Cerenkovcounting.

Assays for cellular and viralgene expression.Cytoplasmic

RNA sampleswereanalyzedtodeterminethe relative

con-tents of c-myc, HIV, and actin RNA. The assays were

performed by dotblotting ofthecytoplasmicRNA samples

onto Schleicher & Schuell, Inc., nitrocellulose filter

mem-brane and hybridization with the appropriate radiolabeled

probe.

Samples for dot blotting were made 25 mM

morpholine-propanesulfonate (pH

7.0)-5

mM sodium acetate-5 mM

EDTA (lx MOPS buffer-S50% deionized formamide-3.7%

formaldehyde

(vol/vol)

in a final volume of 250 ,ul. The

sampleswere incubatedat

56°C

for20min and thenchilled briefly. Nitrocellulose membrane that had been immersed previouslyinwaterand

again

in20x SSCwas

placed

inthe

dot-blotting

manifold;

then

60-,ul

portions

of each

sample

were placed inwells, and theliquid wasdrawnthroughthe

membrane with gentle vacuum. The membrane was then

rinsed oncein2x

SSC,

baked at

80°C

for2 h ina vacuum

oven, and stored until

hybridization.

Radiolabeled probe for

expression

ofthec-myc gene was

generated

by

primer

extension

labeling (11)

of a

1.0-kilobase-pair

(kb)

PstI

fragment

ofthe human c-myc gene

(37). The radielabeledHIV probes were

prepared

by

using

the9.5-kb

fragment

ofthe

plasmid pARV-2 (27),

and actin probes were derived from the 0.6-kb PstI

fragment

of a

murine

,-actin

cDNA clone

(28).

Prehybridization

and

hy-bridization of the membranes were described

previously

(32).

Analysis of proviral genomes. Pellets

containing

infected

cell nucleiwererecovered

subsequent

toNonidet P-40

lysis

and

suspended

in 2.5 ml of25 mM Tris

hydrochloride

(pH

7.5)-12.5 mM EDTA-140 mMNaCl-0.5% sodium

dodecyl

sulfate; proteinase

K was added to afinal concentration of

200

,xg/ml,

and the

samples

were incubated at

56°C

for 30 min.Then,asecond

portion

of

proteinase

Kwas

added,

and

the incubation continued for an additional hour before the

sample was extracted with an

equal

volume of buffered phenol solution. The aqueous

phase

was collected and

dialyzed

extensively against

10 mMTris

hydrochloride

(pH

7.5)-i

mM EDTA. The DNAconcentration was calculated fromthe

optical

density

at260nm.

Purified cellular DNA was

digested

with the restriction

endonuclease

HindIII,

and the

resulting fragments

were

separated by

agarose

gel

electrophoresis.

The DNA was

transferred to

Hybond

filter membrane

(Amersham

Corp.);

hybridization of the filter membrane was

performed

as

describedfor the RNA

analysis.

RESULTS

HIVproduction isrestricted in U937 cells. The

replication

of HIV in U937 cells was evaluated

by

measuring

the

accumulated reverse

transcriptase

activity

and

p24

gag

on November 10, 2019 by guest

http://jvi.asm.org/

C-)

0

Lfl

z

0 2 5

(A)

0-C)

or.

b z ,

ooz C o

~ x

m c>,

z =

L .>

-500 e 2 0

LU4

_ E6

0

z DAYS

FIG. 1. LogarithmicallygrowingU937 cellswere infectedwith 0.2TCID50 per cell of LAVI (passaged in CEMcells).The culture was split, andhalf was treatedwith TPA and washedasdescribed in Materials and Methods.Infected cellswerethenplatedat adensity of 5x 105/ml. Sampleswereremovedimmediately (day0) anddaily for the remaining 5 days and assayed for reverse transcriptase activity (36). Circles denotereversetranscriptaseactivity data,and triangles designate p24antigenmeasurements;opensymbolsdepict the datafor undifferentiated U937cellcultures, and solidsymbols represent TPA-treated cell cultures. The data shown are from a

single experiment (the same culture that was used for RNA and DNAanalyses shownsubsequently) andarehighlyrepresentativeof more than15repetitions.

antigen content of infected-cell supernatants. U937 cells were infected with 0.2TCID50percell; newly synthesized

virus was released intothemedium ondays 1and 2

postin-fection. Thereafter, virus production

decreased

rapidlyand

remained at a low level for the duration of the 5-day experiment (Fig. 1). The decrease in supernatant reverse

transcriptase activity was due toan effecton virus produc-tion and not to a direct inhibition ofreverse transcriptase activity. Admixing experiments, in which virus stocks of

known activity and culture supematants were combined, showed noinhibition of enzymatic activity.

Theeffects of phorbolester treatment onHIVreplication

in U937cellswereevaluatedin asimilarmanner. Cellswere

preinfected ina single flask andthensplit intoTPA-treated and untreated samples. Preinfection was required because

TPA treatment of U937 cells rendered them

resistant

to

infection (7; C.D.P., unpublished data). Infected phorbol ester-treated cells did not release detectable reverse

tran-scriptase orp24 gag antigen (Fig. 1). The results on HIV

production by normal and differentiated U937 cells were

obtained

consistently

in more than 15 independent

experi-ments.

We also assessed the accumulation ofreverse

transcrip-taseactivity inthemedium ofinfectedcells that were treated

with either1,25dihydroxy vitaminD3 ordelectinated

super-natantfromphytohemagglutinin-activated human peripheral

blood lymphocytes. Both treatments are known to induce U937 differentiation (38). In three separate analyses of the

effects of both ofthese treatments, the release of reverse

transcriptaseactivity was similar to that observed with TPA

treatment (Fig. 2). Three

distinct

agents known to induce

U937 cellulardifferentiationthus were able toreduce

signif-icantly the release of HIV from infected cellcultures. The abrupt decline in virus production manifested by

undifferentiated U937 cells was not attributable to cell

tox-icity. Nosignificantdifferences in the rates of DNA

synthe-sis were apparent between the infected and uninfected cell

culturesasdeterminedby the rates of[3H]thymidineuptake

(Fig. 3). The percentage of viable cells as determined by

0 3

DAYS

4 5 6

FIG. 2. Effects ofTPA, 1,25dihydroxy vitamin D3, and PHA-activatedlymphocytesupernatantsontheaccumulation of cell-free reverse transcriptase activity in infected U937 cultures. Infection and cellculturewere asdescribed forFig.1. Thereverse transcrip-taseactivity data arerepresented by theheight of theappropriate bars, and the range of values among threeindependentexperiments is shown by the error bars. Within each group offour bars, the leftmost open bar represents untreatedcells,thehatched bar

desig-natesTPA-treatedcells,the solid bar shows vitaminD3effects

(10-'

Mfinalconcentration), and therightmosthatched bar documents the effect ofadding activated lymphocyte supernatants at 20% final volume.

trypan blue dyeexclusionwasgreater than90% in both the

infectedand uninfected cultures. Somecell death observed

in the phorbol ester-treated cultures was attributable to

chemicaltoxicity.

Viral RNA and provirus accumulation. Levels of HIV

RNAinthe

cytoplasmic

RNA

fraction

were

roughly

corre-lated to the levels of virus

production

in undifferentiated U937 cells. Accumulation of

HIV-specific cytoplasmic

RNA

was observed on days 1 and 2

postinfection;

the RNA decreased in abundanceon

day

5when virusproductionwas

also atits lowest

(Fig.

4). This correlationbetween

cytoplas-mic RNA levelsand virus

production

was notobserved in the differentiated cellcultures. Differentiated

(TPA-treated)

U)

In

0

x

C-c

x

0)

C

E 0.r C)

1500

-1000

-500

-_0_-_

0

0 1 2 3 4 5

[image:3.612.60.291.57.197.2]

Days

FIG. 3. [3H]thymidine uptake in infected and uninfected U937 cell cultures. [3H]thymidine (1 ,uCi) was added to 200 ,ul of cell culture;2.5hlater the cells were recoveredbyfiltration and washed with several volumes of 0.9% NaCl. The data are expressed as countsperminute of[3H]thymidinetaken upin 2.5 hby 105 cellsfor eachday of cell culture. Solid lines representthe uptakerates for uninfected U937cells, andthebroken linesshowthecorresponding ratesfor infected cells. Solid circles denote untreatedcultures, and opencircles mark the cultures treated for 2 h with TPA.

I1,

on November 10, 2019 by guest

http://jvi.asm.org/

[image:3.612.319.546.79.209.2] [image:3.612.315.554.490.650.2]

A

untreated

I

I

is*

.4

TPA-treated

I

I1

*

-ff10.l pg

2 pg 0.5 pg

B if

[0

*b..

infected *-

to

-0

*-

0

10 pg

2 pg 0.5 pg

10 pg *

*

2pg

uninfected

1 2 4 5 1 2 4 5

[image:4.612.67.296.76.307.2]

days post-infection

FIG. 4. Cytoplasmic RNAdot blots showing the levels of HIV (A) and actin(B) RNA in untreated and TPA-treated U937 cells. The blotswerepreparedasdescribed inthetextandhybridized witha radiolabeled9.0-kb DNAfragment ofthe clonedgenomeof the HIV

isolateARV-2 (27)or anEcoRIfragment ofmouse actin.In panel A, only the samples frominfected cells areshown. Therewas no detectable hybridization to uninfectedcell RNA. In panel B, the

uppersectionshows theactinRNAcontentof uninfected cells and the lower panel shows the actin content of infected cells. The

amountoftotalcytoplasmicRNAperdot isshownatthe right, and thedaypostinfection is shownatthe bottom. These RNA samples

werepreparedfromthesamecell cultures represented in Fig.1.No

RNAwasloaded in the day2,uninfected TPA-treated lane, and this

accountsforthe absence ofhybridizationatthis position.

cells contained significant amounts of cytoplasmic viral

RNAeventhough viralreversetranscriptase and p24 antigen

were notreleased into the medium. The level of viral RNA

remained constant in differentiated U937 cells over the

course of the experiment. Hybridization with the mouse

,B-actingeneprobe demonstrated that equalamountsof RNA

had been applied to all positions of the filter and that the

levels of actin RNA were not affected significantly by

infection or differentiation (Fig. 4). Similar results were

obtained inthree additionalexperiments.

Therelative accumulation of HIVprovirus was assessed

by DNA blot analysis (Fig. 5). The hybridization with

samples from days 2.5 and 5 postinfection indicated a

relativelyconstant amountofproviral DNA. In this

partic-ularexperiment, the sample in lanebcontained lessDNA,as

judged by ethidium bromide staining of the gel, and this

probably accountsfor the difference in hybridization

inten-sity between lanes b andc.Therewaslittle difference in the

relative amounts of the provirus between differentiated and

undifferentiated infected cell cultures, and this result was

consistent in three separate infection experiments.

There-fore, neither the decrease in virus production during

re-stricted replication nor the absence of particle release in

differentiated cell cultures could be attributed to selective

depletion of the infected-cell population.

Expression ofthec-mycgeneininfected U937 cells.

Expres-sion of thec-mycgene isamarker ofcellular differentiation

in U937 cells (28). Accordingly, we examined the

accumu-a b c d e

5.0 kb

[image:4.612.379.500.77.136.2]

Vn- 2.5 kb

FIG. 5. Southern blot analysis of HIV proviral DNA in untreated and TPA-treatedU937 cell cultures. Nuclear DNA was prepared as described in the text, digested with the restriction enzyme HindIll, andfractionated on an0.8% agarose gel. After transfer to nitrocel-lulose andhybridization with a probe from the plasmid pARV-2 (27),

the hybridization pattern of the provirus was revealed. Only an

upper section of thehybridization pattern is shown. Lanes contain the following samples: a, uninfected U937 cell DNA; b, day 2, infected, no TPA; c, day 5, infected, no TPA; d, day 2, infected, TPA-treated; and e, day 5, infected, TPA-treated. The sizes of the bandsareshownatthe right.

lated levels of cytoplasmic c-myc RNA in infected and

uninfected cells, either with or without exposure to TPA

(Fig. 6). The dot blots wereprepared by using the same RNA

samples shown in Fig. 4, thus direct, qualitative comparison

betweenrelative levels of HIV and c-myc RNA is possible.

A nearly constant level of c-myc RNA was observed in

undifferentiated cells; the relative amountdeclined rapidly

after TPA treatment. This finding is in agreement with the

results reportedpreviouslyby Mitchell et al. (28) and attests

to therelationshipbetween c-myc RNA levels and the state

of monocyte differentiation. Infection by HIV altered

signif-icantly the patternof c-myc gene expression in U937 cells.

Instead ofremaining constant in the undifferentiated cell

cultures,c-myc RNA declinedsteadilyand inparallelto the

decreasing level of HIV RNA in these infected but still

viable cells. It is important to note that the decrease in

relative abundanceofc-myc RNA shown in Fig. 6 is most

likelyanunderestimate of the actualmagnitudeof change in individual infected cells because the RNA samples in these

experiments wereprepared from cell cultures composed of

approximately equal numbers of infected and uninfected

cells as determined by indirect immunofluorescence (not

shown). Therefore, the apparent change in c-myc RNA

probably reflects a larger decline in abundance of this

untreated

TPA-treated

*s,*. *gm1o10

pg

infected

L

*

-

2

pg

**@S@.*OO2*g0.5,ug

@0

0---

e

4

10

Pg

uninfected

*

**

a

_{2 pg}

L .

,

0.5

pg

1 2 4

5

1 2 4 5

days post-infection

FIG. 6. Relativecontentof c-mycRNAin infected and uninfec-ted U937 cells withandwithout TPAtreatment.Theupper section shows the levels of c-myc RNA in infected cells, and the lower sectionshows thelevelsinuninfected cells. The RNAsamplesare

identical to thoseused in Fig. 4, thus the ,-actin control can be

compared directlywith thec-mychybridizationpattern.

on November 10, 2019 by guest

http://jvi.asm.org/

[image:4.612.319.561.519.672.2]

species

in infected cells

superimposed

on the

normally

constant RNA levels in uninfected cells. We have not yet

examined the

pattern

of c-myc gene

expression

in cell

cultures

containing

100% infected cells because at

higher

multiplicities

ofinfection the property of restricted

replica-tion isnotobserved

(C.

D.

Pauza,

unpublished).

Incontrast tothe undifferentiatedinfectedU937

cells,

the abundance of

c-myc RNA remained

nearly

constant in the infected

TPA-treatedcells

throughout

the

5-day

timecourse. Inthiscase,

the altered

regulation

ofc-myc RNAaccumulation is

espe-cially

apparent. The uninfected differentiated cell cultures

would

normally

be

expected

to contain little or no c-myc

RNA after 2

days

of TPA treatment.

Thus,

the relative

abundance ofc-myc RNA in theinfecteddifferentiated cells was

clearly

evident.

On the basis of these

experimental observations,

we

concludethat HIV infection altered the

regulation

ofc-myc RNA accumulation.

Surprisingly,

a correlation between

HIV and c-myc RNA levelswas observed. This is different

from thepatternofc-myc

expression

observed inuninfected

cells. Altered

expression

ofthe c-myc genewasobservedin

four separate infection

experiments;

in each case, the

changes

in c-myc RNA were

qualitatively

similar to the

patterns ofviral RNA accumulation. Thisresult appears to

be a

specific

effect ofviral infection because

expression

of the actin genewasnotaffected

by

HIVinfection.

DISCUSSION

HIV

replication

in U937 cells is

productive initially

and

thenbecomes restricted.

During restriction,

thelevelofviral RNA in the

cytoplasm

of infected cells decreases

substan-tially.

Expression

ofthe c-myc genealso decreases

concom-itantly

with viral gene

expression.

DNA blot

analysis

showedthat thesameamountsof

proviral

DNAper cellare

present

during

both

productive

and restricted

replication;

thisobservation attests tothefactthat the

changes

in RNA

levels indicate altered accumulation of these

species

in

individual cells andnotselection

against

a

subpopulation

of infected cells.

Furthermore, thymidine uptake

rates for

infected and uninfected cells are

comparable, indicating

furtherthat thealteredbehavior of infectedcellsisnotdueto

a

cytopathic

effect ofthe virus.

Consequently,

the

stringent

control of HIV

replication

in U937 cells is an intrinsic attribute of the interaction between the monoblastoid cell

and the virus.

Infected U937

cells,

induced to differentiate

by phorbol

ester, release reduced amounts ofHIV. In contrast to the

restricted

replication

observed in undifferentiated U937

cells,

the differentiated cellswere

nonproductively

infected

despite

accumulation of

significant

levels of

cytoplasmic

viralRNA. Inthiscase,

high

levels ofc-myc RNA werealso

showntobe present

during

differentiationof infected cells in

contrast to the

rapid

decline in abundance of this RNA

observed in differentiated uninfected cells. Infected U937

cells were also induced to differentiate

by

the addition of

1,25

dihydroxy

vitamin

D3

(4, 5)ordelectinatedsupernatant

derived from PHA-activated human peripheral blood

lym-phocytes.

In all three cases, the infection was rendered

nonproductive

in the treatedcells.Differentiationof infected

U937cellsinitiated

by

anyofthree

independent

compounds

resulted inan

abrupt

decline in virusproduction.

The interaction between HIV and the host cell U937 is

complex.

Undifferentiated

cells,

afteraninitialburst of virus

production

in thefirst few

days

afterinfection,subsequently

show restricted virus

replication.

Incontrast, theinfectionof

differentiated cells is nonproductive despite high levels of

viral RNA accumulation. In both cases,provirus

accumula-tion isrelativelyconstant.Alterations in the pattern of c-myc

but not actin gene expression reveal that HIV infection

exerts aspecific and selective effect on cellular gene

expres-sion. These results demonstrate that the mechanisms

con-trollingHIVreplicationin monoblastoid cellsarespecificto

the cellular differentiation state. Inaddition, virus infection

itself resulted inphenotypicchangesin the monoblastoid cell

populationwhich suggest that HIVcanalter thecapacity of

these cellsto differentiate.

Others have noted that infected U937 cells manifest

al-teredphenotypes and that their differentiation is somewhat

inhibited (7, 21). Here we present molecular evidence in

supportof this concept and introduce the findingthat virus

production is diminished in normal and differentiated U937

cells; the operative mechanisms in these two cases are

functionally distinct.

Itis of interest to compare our results with the

observa-tions of Folks et al. (12, 13)onthe behavior ofa

persistently

infected cloned cell line (designated Ul) derived from the

parent line U937. The Ul line showed minimal constitutive

expression of HIV, and treatment of these cells with

phy-tohemagglutinin-activated lymphocytes supernatants

in-creased virus production (12). In ourhands,this treatment

inevitably causes a rapid decline in virus release, which is

consistent with the observed correlation between cellular

differentiationand reduced HIVproduction. The differences

between our observations on thepattern of HIVreplication

following acute infection of the U937 cell line and those of

Folks et al. (13) on the chronically infected Ul cell line

remain to beelucidated.

HIV replication in differentiated and undifferentiated

U937 cells is very similar to the pattern of visna virus

replication in normal sheep monocytes and macrophages. In

the visna virus studies, undifferentiated monocytes were

permissive and showed low levels of virus replication.

Differentiation oftheinfectedmonocytes wasinduced by an

interferon activity (16, 24, 29) and was accompanied by

increased viralgeneexpression (16, 29). Despite the elevated

levels ofcytoplasmic viral RNA, infectious virus was not

released from the differentiated cells (16). Thus, the same

relationshipbetweendifferentiationandviralreplication that was documentedhere for HIV in human monoblastoid cells

was observed previously for another lentivirus infection of

normal sheep monocytes and macrophages. Visna virus

infection ofsheep macrophages is associated with the

for-mation ofareservoir ofvirus (24) and the establishment of a

persistent infection(16,17).Thesetwo characteristics, virus

persistence and the establishment of a virus reservoir, are also important components in HIV infection. The state of

human monocyte differentiation and its impact on virus

replicationislikely to be central to the mechanisms

control-ling these two features of HIV infection and of crucial

importance in thepathogenesis of AIDS.

ACKNOWLEDGMENTS

C.D.P. wassupported by Public Health Service grants AI-21243 and AI-05875 and Cancer Training Grant CA 09254 from the National Institutes ofHealth (to Melvin Cohn) and by grant 000557 from the American Foundation for AIDS Research. D.D.R. was supported by the VeteransAdministration grant HL-32471 and by PublicHealthService contracts HB-67019 from the National Heart, Lung,andBloodInstitute andAI-52578from the National Institute ofAllergyand InfectiousDiseases.

on November 10, 2019 by guest

http://jvi.asm.org/

LITERATURECITED

1. Alizon, M., P. Sonigo, F. Barre-Sinousi, J.-C. Chermann, P. Tiollais, L. Montagnier, and S. Wain-Hobson. 1984. Molecular cloning oflymphadenopathy-associated virus. Nature (London) 312:757-760.

2. Armstrong, J. A., R. L. Dawkins, and R. Horne. 1985. Retroviral infection of the accessory cells and the immunological paradox in AIDS. Immunol. Today 6:121-122.

3. Armstrong, J. A., and R. Horne. 1984. Follicular dendritic cells and virus-like particles in AIDS-related lymphadenopathy.

Lan-ceti:370-372.

4. Bar-Shavit, Z., S. L. Teitelbaum, P. Reitsma, A. Hall, L. E. Pegg, J. Trial, and A. J. Kahn. 1983. Induction of monocytic differentiation and bone resorption by 1,25-dihydroxyvitamin

D3. Proc. Natl. Acad. Sci. USA 80:5907-5911.

5. Bar-Shavit, Z., S. L. Teitelbaum, G. P. Stricklin, A. Z. Eisen, A. J. Kahn, and H. G. Welgus. 1985. Differentiation of a human leukemia cell line and expression of collagenase inhibitor. Proc. Natl. Acad. Sci. USA 82:5380-5384.

6. Chayt, K., M. Harper, L. Marselle, E. Lewin, R. Rose, J. Oleske, L. Epstein, and R. Gallo. 1986. Detection ofHTLV-III RNA in lungs of patients with AIDS and pulmonary involvement. J. Am. Med. Assoc. 256:2356-2359.

7. Clapham, P., R. Weiss, A. Delgleish, M. Exley, D. Whitby, and N. Hogg. 1987. Human immunodeficiency virus infection of monocytic and T-lymphocytic cells: receptor modulation and differentiation induced by phorbol ester. Virology 158:44-51. 8. Cohn, Z. A., and B. Benson. 1965. The differentiation of

mononuclear phagocytes. Morphology, cytochemistry, and bio-chemistry. Int. J. Exp. Med. 121:153-169.

9. Estevez, M. E.,I.J. Ballart, R. A. Diez, N. Planes, C. Scaglione, and L. Sen. 1986. Early defect of phagocytic cell function in

subjects at risk for acquired immunodeficiency syndrome. Scand. J. Immunol. 24:215-221.

10. Favoloro, J., R. Treisman, and R. Kamen. 1980. Transcription maps of polyoma-virus specific RNA: analysis by two-dimen-sional nuclease S1 gel mapping. Methods Enzymol. 65:718-749.

11. Feinberg, A. P., and B. Vogelstein. 1983. A technique for radio labeling DNA restriction endonuclease fragments to high spe-cific activity. Analytic Biochem. 132:6-13.

12. Folks, T., J. Justement, A. Kinter, C. Dinarello, and A. Fauci. 1987. Cytokine-induced expression of HIV-1 in a chronically infected promonocyte cell line. Science 238:800-820.

13. Folks, T., J. Justement, A. Kinter, S. Schnittman, J. Orenstein, G.Poli, and A. Fauci. 1988. Characterization of a promonocyte clone chronically infected with HIV and inducible by 13-phorbol-12-myristate acetate. J. Immunol. 140:1117-1122. 14. Gabuzda, D., D. Ho, S. de la Monte, M. Hirsch, T. Rota, and R.

Sobel. 1986. Immunohistochemical identification of HTLV-III antigens in brains of patients with AIDS. Ann. Neurol. 20:289-291.

15. Gartner, S., P. Markovits, D. M. Markovitz, M. H. Kaplan, R. C. Gallo, and M. Popovic. 1986. The role of mononuclear phagocytes in HTLV-III/LAV infection. Science 233:215-219.

16. Gendelman, H. E., 0. Narayan, S. Kennedy-Stoskopf, P. G. E. Kennedy, Z. Ghotbi, J. E. Clements, J. Sta;dley,and G. Pezeshk-pour. 1986. Tropism of sheep lentiviruses for monocytes: sus-ceptibility to infection and virus gene expression increase during maturation of monocytes to macrophages. J. Virol. 58:67-74.

17. Gendelman, H. E., 0.Narayan, S. Molineaux, J. E. Clements, and Z. Ghotbi. 1985. Slow, persistent replication of lentiviruses: role of tissue macrophages and macrophage precursors in bone marrow. Proc. Natl. Acad. Sci. USA82:7086-7090.

18. Gordon, S., J. Todd, and Z. A. Cohn. 1974. In vitro synthesis and secretion oflysozymeby mononuclear phagocytes. J. Exp. Med. 139:1228-1248.

19. Grieco, M. H., M. H. Reddy, H. G. Kothari, M. Lange, E. Buimovici-Klein, and D. William. 1984. Elevated P2-micro-globulin and lysozyme levels in patients with acquired immune

deficiency syndrome. Clin. Immunol. Immunopathol. 32:174-184.

20. Haertle, T., C. Carrera, J. S. McDougal, L. Sowers, D. D. Richman, and D. A. Carson. 1988. Metabolism and anti-HIV activity of 2'-halo-2',3'-dideoxyadenosine derivatives. J. Biol. Chem. 263:5870-5875.

21. Hammer, S., J. Gillis, J. Groopman, and R. Rose. 1986. Invitro modification of human immunodeficiency virus infection by granulocyte-macrophage colony-stimulating factor and gamma interferon. Proc. Natl. Acad. Sci. USA 83:8734-8738.

22. Ho, D., R. Pomerantz, and J. Kaplan. 1987. Pathogenesis of infectionwithhuman immunodeficiencyvirus. N.Engl.J.Med. 317:278-286.

23. Ho, D. D., T. R. Rota, and M. S. Hirsch. 1986. Infection of human monocyte/macrophages by human T-lymphotropic virus typeIII. J. Clin. Invest. 77:1712-1715.

24. Kennedy, P. G. E., 0. Narayan, Z. Ghotbi, J. Hopkins, H. E. Gendelman, and J. E. Clements. 1985. Persistent expressionof Ia antigen and viral genome in visna-maedi virus-induced in-flammatorycells. J. Exp. Med. 162:1970-1982.

25. Klatzmann, D., and J. C. Gluckman. 1986. HIVinfection: facts and hypotheses. Immunol. Today 7:291-296.

26. Koenig, S., H. E. Gendelman, J. M. Orenstein, M. C. dal Canto, G. H. Pezeshkpour, M. Yungbluth, F. Janotta, and A. Aksamit. 1986. Detection of AIDS virus in macrophages in brain tissue from AIDS patients with encephalopathy. Science 233:1089-1093.

27. Luciw, P. A., S. J. Potter, K. Steimer, D. Dina, and J. A. Levy. 1984. Molecular cloning of the AIDS-associated retrovirus. Nature(London) 312:760-763.

28. Mitchell, R. L., L. Zokas, R. D. Schreiber, and I. M. Verma. 1985. Rapid induction ofthe expression ofproto-oncogenefos during human monocytic differentiation. Cell 40:209-217. 29. Narayan, O., D. Sheffer, J. E. Clements, and G. Tennekoon.

1985.Restricted replication oflentiviruses: Visna virusesinduce aunique interferon during interaction betweenlymphocytes and macrophages. J. Exp. Med. 162:1954-1969.

30. Nicholson, J. K. A., G. D. Cross, C. S. Callaway, and J. S. McDougal. 1986. In vitro infection of human monocytes with human T lymphotropic virus type III/lymphadenopathy-associ-ated virus (HTLV-III/LAV). J. Immunol. 137:323-329. 31. Pallesen, G., J. Gerstoft, and L. Mathiesen. 1987. Stages of

LAV/HTLV-III lymphadenitis.I. Histological and immunolog-ical classification. Scand. J. Immunol. 25:83-91.

32. Pauza, C. D. 1987. Regulation of human T-lymphocyte gene expressionby interleukin-2: immediate-response genes include the proto-oncogene c-myc. Mol. Cell. Biol. 7:342-348. 33. Pauza, C. D. 1988. HIV persistence in monocytes leads to

pathogenesis andAIDS. Cell. Immunol. 112:414-424.

34. Plata, F., B. Autran, L. Martins, S.Wain-Hobson, M.Raphael, C. Mayaud, M. Denis, and J. Guillon. 1987.AIDSvirus-specific cytotoxic T lymphocytes in lung disorders. Nature (London) 328:348-351.

35. Poli, G., B. Bottazzi, R. Acero, L. Bersane, V. Rossi, M.Introna, A. Lazzarin, and A. Mantovani. 1985. Monocyte function in intravenous drug abusers withlymphadenopathy syndrome and in patients with acquiredimmunodeficiency syndrome: selective impairment of chemotaxis. Clin. Exp. Immunol. 62:136-147. 36. Popovic, M., M. G. Sarngadharan, E. Read, and R. C. Gallo.

1984. Detection, isolation, and continuous production of cyto-pathic retroviruses (HTLV-III) from patients with AIDS and pre-AIDS. Science 224:497-500.

37. Rabbitts, T. H., P. H. Hamlyn, and R. Baer. 1983. Altered nucleotide sequences of atranslocated c-mycgene in Burkitt's lymphoma. Nature (London)306:760-765.

38. Ralph, P. 1981. Continuous macrophage cell lines-their use in the study of induced and constitutive macrophage properties and cytotoxicity, p. 175-195. In M. A. Landy. Lymphokines, vol. 4. Academic Press, Inc., Orlando, Fla.

39. Richman, D., R. Kornbluth, and D. Carson. 1987. Failure of dideoxynucleosides to inhibit human immunodeficiency virus replication in cultured humanmacrophages. J. Exp. Med. 166: 1144-1149.

on November 10, 2019 by guest

http://jvi.asm.org/

40. Smith, P., K. Ohura, H. Masur, H. Lane,A.Fauci, andS.Wahl. 1984. Monocyte function in the acquired immune deficiency syndrome: defective chemotaxis. J. Clin. Invest. 74:2121-2128.

41. Sundstrom, C., and K. Nilsson. 1976. Establishmentand char-acterization ofahumanhistiocytic lymphoma cell line (U937).

Int. J. Cancer17:565-577.

42. Weiss, A., H. Hollander, and J.Stobo. 1985. Acquired immuno-deficiency syndrome: epidemiology,virology andimmunology. Annu. Rev. Med.36:545-562.

43. Wiley, C., R. Schrier, J. Nelson, P. Lampert, and M.Oldstone. 1986. Cellular localization of human immunodeficiency virus infectionwithin the brains ofacquiredimmunedeficiency

syn-dromepatients. Proc. Natl. Acad. Sci. USA83:7089-7093.

on November 10, 2019 by guest

http://jvi.asm.org/