Tumor necrosis factor-alpha modulates
monocyte/macrophage apoprotein E gene
expression.
H Duan, … , Z Li, T Mazzone
J Clin Invest.
1995;
96(2)
:915-922.
https://doi.org/10.1172/JCI118139
.
apo E has been shown to modulate cholesterol balance in arterial wall cells. Production of
apo E by macrophages in atherosclerotic plaques could thereby influence the development
of the plaque lesion. Cytokines, including TNF alpha, have been identified in human
lesions, therefore, we undertook a series of studies to evaluate the effect of TNF alpha on
monocyte/macrophage apo E production. The addition of TNF alpha to freshly isolated
human monocytes led to a four- to fivefold increase of apo E mRNA abundance. The
addition of TNF alpha to fully differentiated macrophages either had no effect or modestly
inhibited apo E mRNA expression. THP1 human monocytic cells also responded to TNF
alpha in a phenotype-specific manner. Treatment of these cells with TNF alpha produced a
dose- and time-dependent increase in apo E mRNA. This increase was reflected in apo E
synthesis and was associated with inhibition of DNA synthesis, and with induction of c-fos
and ICAM-1 gene expression. Cell-permanent analogues of ceramide did not reproduce
TNF alpha effect on apo E, but antagonists of protein kinase C did inhibit its effect. TNF
alpha induction of apo E mRNA abundance was associated with stimulation of apo E
promoter-dependent gene transcription. In summary, TNF alpha stimulates apo E gene
transcription, mRNA abundance, and protein synthesis in the monocyte/macrophage in a
phenotype-specific manner. Such […]
Research Article
Tumor Necrosis Factor-
etModulates Monocyte/Macrophage
Apoprotein E
Gene
Expression
Hongwei Duan, Zhigao Li, and Theodore Mazzone
Departmentsof Medicine and Biochemistry, Rush Medical College, Chicago, Illinois 60612
Abstract
apo E has been shown to modulate cholesterol balance in arterial wall cells.Production of apo E by macrophages in atheroscleroticplaques could thereby influence the
develop-mentoftheplaque lesion.Cytokines, includingTNFa, have beenidentifiedin humanlesions, therefore,weundertooka
series of studiestoevaluate theeffect ofTNFaonmonocyte/ macrophage apo E production. The addition of TNFa to
freshly isolatedhuman monocytes ledto afour-to fivefold increaseof apo E mRNA abundance. The addition of TNFa to fully differentiatedmacrophages either had noeffect or modestly inhibited apo E mRNAexpression. THP1 human monocytic cells also responded to TNFa in a phenotype-specific manner. Treatmentofthese cells with TNFx pro-ducedadose- andtime-dependentincrease in apo E mRNA. This increase wasreflected in apo E synthesis and was asso-ciated with inhibition of DNAsynthesis,andwith induction ofc-fosand ICAM-1 geneexpression. Cell-permanent
ana-logues ofceramide did not reproduce TNFa effectonapo E, butantagonists ofprotein kinase C did inhibit itseffect. TNFa induction of apo E mRNA abundance was associated with stimulation ofapo E promoter-dependent gene tran-scription.In summary, TNFa stimulatesapo E gene
tran-scription, mRNA abundance, and protein synthesis in the monocyte/macrophage in a
phenotype-specific
manner.Such regulation could significantly modify the amount of apo E present in vessel wall lesions. (J. Clin.Invest. 1995. 96:915-922.) Keywords:cytokines * atherosclerosis *
mac-rophage activation *proteinkinase C *ceramide
Introduction
apo E is an importantstructural and functional constituent of plasma lipoproteins, including VLDL, HDL, andchylomicron
remnants (1). Although apo E is primarily synthesized by the liver, significant amounts are also produced by a variety of peripheraltissues and cells (2), including macrophages, where its synthesis is regulated by the cholesterol content of the cells (3, 4).Histochemical studies have indicated that macrophages
Address correspondence to Theodore Mazzone, M.D., Rush Medical College, 1653 W.Congress Parkway, Chicago, IL60612. Phone: 312-942-8231; FAX: 312-563-2096.
Receivedfor publication24October 1994 andacceptedinrevised form19April1995.
are the major source of locally derived apo E in developing vessel wall atheromatous lesions (5). Local productionof apoE inthe vessel wall could impact on proximate cellular cholesterol flux and modify the local availability of bioactive molecules such as growth factors (1). Furthermore, endogenous produc-tion of apo E has been showntoinfluencemacrophage choles-terolflux byregulating the rate of cellular cholesterol clearance to HDL3(6). Thus,modulation of macrophage apo E secretion in the vessel wall couldsignificantly impact atheroma progres-sionorregression.
Regulation of macrophage apo E expression is complex: it isregulated not only by cholesterol but also by activators of protein kinase C and by endotoxin (7-9).
TGF,8
has been demonstratedtoinducemacrophage apoEexpression (10). On the other hand, the monocyte colony stimulating factor(MSCF)'
has beenshown, in separate reports, bothto inhibit and augmentmacrophage apo E expression ( 10, 11). A poten-tial regulatory role for othercytokines, for example, TNFa, has beenevenless clear(10, 12).Local production of TNFa is animportant component of the vessel wall response to injury. TNFa positivity has been found in thecytoplasm of macrophages, and in the cytoplasm and attached to cell membranes of smooth muscle cells and endothelial cells of the human atheroma ( 13, 14). These obser-vations suggest a potential for TNFa to modulate the evolution of the atherosclerotic plaque. Further, such modulation would be likely to occur via effects on monocyte/macrophage cells because theseareprimary targets of TNFa action. In the present study,weinvestigated the effect of TNFa on apoE expression inthe humanmonocyte/macrophage. We examined phenotype-dependent effects of TNFaonapo E geneexpression and
com-pared these with its effectsontheexpression of other monocyte/ macrophage genes (c-fos and intercellular adhesion molecule [ICAM] -1 ) and cell cycle traverse. We also investigated poten-tialroles for ceramide second messengers and protein kinase C (pkC) activation for transducing the observed effects of TNFa
onapo E.
Methods
Materials. [35S]Methionine (10 Ci/mmol), [32P]dCTP (800 Ci mmol), and [3H]thymidine (5Ci/mmol)werepurchased from Amer-shamCorp. (Arlington Heights, IL). Human recombinant TNFa was purchased from Genzyme, Corp. (Boston, MA).C2-ceramide and H7
werepurchased from BIOMOL Research Laboratories, Inc. (Plymouth Meeting, PA). Endotoxin-free BSAcamefrom CalbiochemCorp. (La
1.Abbreviations used in thispaper:ICAM,intercellular adhesion
mole-cule; MSCF,monocytecolony stimulating factor; NFKB,nuclease
fac-tor-KB; pkC, proteinkinase C.
J.Clin. Invest.
()The American Society for Clinical Investigation, Inc. 0021-9738/95/08/0915/08 $2.00
Jolla, CA). All othermaterialswere obtained from previously identified sources (6, 9, 15, 16).
Cell culture. Freshly isolated human monocytes were purified by elutriation. The cell population used for experimentswas >95% mono-cytic, as determined by differential counts ofWright stained smears. These cells were incubated using conditions described in the figure legends. For experiments using fully differentiated human monocyte-derived macrophages,cells wereallowedtodifferentiatein serum-con-taining medium for7d before thestart ofthe incubationwithTNFa. The THPl human monocytic cell line was obtained from American TypeCultureCollection(ATCC, Rockville,MD),andcells were grown in culture flasks (Falcon Plastics,Cockeysville, MD) in RPMI-1640 mediumcontaining 10% heat-inactivated FBS asdescribed (15). For experiments, all cells were washed twice with serum-free RPMI-1640 andwereplacedinRPMI-1640 mediumsupplemented with 0.2% endo-toxin-free BSA unless otherwise indicated.
Immunoprecipitationofapo E.THP1 cellswerepulse-labeledwith L-[35S]methionine,asdescribed in detail previously (16). Briefly,cells wereplated at a densityof5 x 106cells in a 60-mm dish in triplicate. 5 ng/ml ofTNFa wasadded, andthecellswereincubated for72 h at 37TC. After incubation, adherent cells and cells in suspension were washedtwice with methionine-free RPMI-1640 and then incubated for 2 h in 1 ml ofmethionine-free RPMI-1640 medium supplemented with 100,uCi/ml of[35S]methionine supplemented with 10 uMunlabeled L-methionine.Afterincubation,themediumwascollected after centrifu-gationto removecells. Adherent and cellsinsuspension werewashed twice inice-cold phosphate-buffered salineandlysed in 2.0% SDSat
950Cfor3 min.Thelysate wasthendiluted withlysis buffer (10 mM Na2HPO4, 15 mMNaCl, 10mML-methionine, 1%TritonX-100, 1% deoxycholate, pH 7.4)tofinalconcentration of 0.2% SDS. Cellextracts and culture media wereanalyzed byimmunoprecipitation followed by 10% SDS-PAGE and fluorography as previously described (16). 4 x 106 TCA-precipitable counts/min were immunoprecipitated from each celllysate sample,and 4x 105 counts/minwere immunoprecipi-tated from eachmedium sample. Incorporation of labeled methionine intoapo EproteininTHP1 cellswaslinearoverthe assayperiod.For quantitation, apo Ebands were excisedfrom thegel and digested in 30%H202for scintillation counting (16). Radioactivity inthe apo E shown wasnormalized forrateoftotalprotein synthesis and secretion, whichwas notsignificantly differentbetween groups.
RNAisolation and Northernhybridization. Cellswereseededat a density of 10 x 106cells in a 100-mmdish. To isolatetotalcellular RNA,washedcells weresolubilizedinguanidine isothiocyanateas de-scribed, followed by sedimentation of theextractthrough CsCl (9).For some experiments, poly (A)+ RNA was purified after passing total RNA through silanized oligo (dT) columns. For Northern analysis,
formaldehyde-treatedRNAsampleswerefractionated by electrophore-sis in 1.0%agarose,transferredtoNytran(Schleicher & Schuell, Inc., Keene, NH), and hybridized with a labeled apo E cDNA probe as
previouslydescribed(9). Radiolabeledapo E cDNAprobewasprepared by random-primed synthesis using pHEA4, which is a 1.1-kb
Aatll-HinfI fragment that encompasses theentire protein coding region of the human apo E cDNAsubcloned intopUC (9). Each reaction was
carried outusing50 ng of cDNA withaminimum of 1 x 107cpm.To
providean internalcontrol,Nytran membranes werestripped according
to themanufacturer's instructions andreprobedwitha labeled 0.7-kb PstIfragmentof the cDNA for beta-actin. Thec-fos probe,which was a 1.0-kbPstIsubfragmentfromthefoscDNA and theICAM-1 probe, which was a 1.4-kb BgIII-SaII subfragmentfrom theICAM-1 cDNA, waslabeled using the same method.
Measurementof DNA synthesis. Rates of DNA synthesis were as-sessedby determining[3H]thymidine incorporation into TCA precipita-blematerial (17). Aftercultures weretreatedwith theindicatedamount of TNFa for 24h,cells werepulsedwith 2.5 0Ci/ml [3H]thymidine
for 2 h and then washed withRPMI-1640fivetimes. This was followed byfive washes with ice-cold 5% TCA. Cellswere thenextracted in 1
a a
I
p4.
;a
apoE
JP-actin
TNFa
:~
iWk 77
00 4W%O
control
TNFaFigure 1. TNFa enhances apo E mRNA expression in human monocyte/ macrophages. Human monocytes (isolated byelutriation, > 95% pure) were incubated in Teflon FEP culture bottles(Fischer, Pittsburgh, PA) in 0.2% endotoxin-free BSA plus or minus TNFa at 10 ng/ml. After 24 h, cells were harvested for analysis by Northern hybridization. The intensity of the bands for apo E and
/l-actin
(as the internal control) was determined by scanning densitometry.ml of 0.25 N NaOH for scintillation counting. Protein was assayed using Lowry's method (18).
Celltransfection and measurement of luciferase activity. An apo E-luciferase chimeric construct, -2300/ +24 apo E pGL2, was made using standard recombinanttechniques (19). The Sma 1-Aatll fragment of the apo E gene was cloned intothepolylikerregion of pGL-2 Basic (PromegaCorp. Madison, WI). THP1 cells were transfected using a modified dextran method (20). 1 x 107 cells were transfected with 2
pgof -2300/+24apo EpGL2 plasmid in1 mlSTBS buffer (25mM Tris-Cl, 137 mMNaCl,5 mMKCl,0.6 mMNa2HPO4,0.7 mMCaCl2, 0.5mMMgCl2) with 100pg DEAEdextran for 90 min and shocked with 10% DMSO for5 min.Aftertransfection, the cellswereincubated inRPMImediumwith10% FBSovernight.Atthat time,alltransfected cellsweremixed in asingle pool priortobeing aliquotedtoexperimental treatments, toeliminate differential transfectionefficiencyas an experi-mental variable. Themedium wasthen changedto RPMIwith 0.2% endotoxin-freeBSA,and 10ng/mlof TNF-a wasadded. Before harvest-ing, the cells wereleftat roomtemperaturefor20 min; the cells were thenlysed,andluciferase activitywasmeasuredusingluciferase assay kit and a luminometer(Enhanced Luciferase AssayKit; Luminometer Monolight 2010; Analytical Luminescence Laboratory, San Diego, CA). Protein concentrationwasmeasuredusingaDCproteinassay kit (Bio Rad Laboratories, Richmond, CA).
Results
Table L Effect of TNFa on apo E mRNA Expression inDifferentiated Macrophages
Celltype apoE/j3-actin Fold change
THPl cells
24 h control 2.58 1.0
24 h +TNFa 1.97 0.76
72 h control 2.46 1.0
72 h +TNFa 1.57 0.64
Humanmonocyte-derived macrophages
48 h control 3.88±0.14 1.0
48 h +TNFa 3.57±0.40 0.92
THPl cells in10%serum were treated with 100 ng/mlofPMAfor 3 d. At that time themediumwasreplacedwith 0.2%endotoxin-freeBSA alone or with5 ng/mlofTNFafortheindicated time. Human monocytes isolated byelutriation were allowed to differentiate in serum containing mediumfor 7 d. At thattime, the medium wasreplacedwith 0.2% endotoxin-freeBSA alone or with 5 ng/ml TNFafor48 h.Cellswere analyzed by Northern blot byhybridization. The intensityofthe bands for apo E and/-actinwasmeasured by scanningdensitometry.Results forTHP1 cells are the average ofduplicate samples,which varied
< 10%. Results from human monocyte-derived macrophage are mean±SD of triplicate samples.
mlafter a 48-h incubation period (Table I). Addition of TNFa
to cells allowed to differentiate over shorter periods of time produced a 30-40% reduction of apo E mRNA levels (not shown). Similar phenotype-specific effects were observed in thehumanmonocytic THP1 cell line. Inthis cell line, weand others have shown that treatment of THP1 monocytes with phorbolesterinducedexpressionof the macrophage phenotype inthese cells as well as in apo E gene expression (8, 9, 12).
When TNFa wasaddedduring the initial 18 h of differentiation, apo E mRNA abundancewasfurther increased more than two-foldby the addition ofTNFa(Fig.2). When TNFa was added
toTHP1 cellsafter differentiation had progressed for 72 h, apo
PMA
PMA+TNFa
III
0
PUA+TNF-a *
I I I
1 2 3 4 5 6 apoE/beta-actin
Figure 2. TNFa enhances apo E mRNA expression in early differentiat-ing macrophages.THP1 cells in0.2% endotoxin-free BSA were treated with 100ng/ml of PMA alone or with 5 ng/ml of TNFa for 18 h. Northern blots of poly (A)+RNAfor apo E and/3-actinwere processed
asdescribed in Methods. The bars show the change of apo E mRNA abundance aftercorrectingfor the/3-actinsignal.
Table H. Dose-dependent Effect of TNFa on apoE mRNA Expression
apoE/f3-actin Foldchange
Control 0.09 1
TNFa5 ng/ml 0.32 3.6
TNFa 10 ng/ml 0.40 4.4
TNFa 15 ng/ml 0.42 4.7
20
qg
total cellular RNA, isolated fromuntreatedcellsorcellstreated with the indicatedconcentrationof TNFa for 24h,wasanalyzed by Northernhybridization. Theintensityof the bandscorrespondingto apo E and/3-actinwasdeterminedbyscanningdensitometry.The values shown are the averagesofduplicate samples, which varied < 10%.E mRNA abundance was reduced by 24 and 36% after an
additional 24 and72h,respectively, inTNFa (Table I). The resultsin both cell types,therefore, demonstratedarelationship betweenTNFaeffect onapoEexpression and phenotypic
ex-pression of monocyte/macrophage cells. Additional studies
wereperformedtoinvestigate further TNFa-mediatedinduction
of apoE gene expressionusing the THP1 cell line.
TNFa induces apo E gene expression in THPI cells in a time- anddose-dependent manner. TableII shows the results ofarepresentativeexperimentto evaluate the effect ofTNFa onapoEgeneexpressionover arange of doses. In this experi-ment, apo E messenger RNA abundance increased 3.6-fold in response to5 ng/ml TNFa and 4.7-fold in responseto 15ng/ mlTNFa. The5 ng/ml dose was chosentoevaluate the time dependency of apoEmRNA response. As shown in Table III, aftera24-h incubation inTNFa,aslightlygreater than twofold increase in apo E mRNA abundance was detected. The greatest increase of apoEmRNAwasobserved when the incubation in TNFa wasextendedto72h, atwhich timea 14-fold increase in apo E messenger RNA abundancewas measured intreated
ascompared with control cells.
Effect
of TNFaon apoEsynthesis andsecretion. Induction of messenger RNA levels for macrophage target genes of TNFaare not always reflected in changes in protein synthesis and secretion (21, 22). We therefore directly evaluated the effect ofTNFa onapo Esynthesis and secretion by pulse labeling cells
Table III. Time CourseEffect of TNFa onapo E mRNA Expression
apoE/,B-actin
Time Control TNFa
18 h 0.09 0.11
24 h 0.07 0.16
72 h 0.09 1.26
20
Mg
of total cellularRNAisolated from untreated cells and cells treated with 5ng/mlofTNFafor the indicated timewasanalyzed by Northernhybridization.Theintensityof the bands correspondingtoTable IV. Effect of TNFa on apo E Synthesis and Secretion
Medium Cell extract
dpm
Control 265±73 447±69
TNFa 1163±49 822±178
5 x106THP1cells in60-mmdishes were incubated for 72 h inmedium alone orwith 5ng/mlofTNFa. The cells were then labeledfor2 h with[35S]methioninetomeasure the rateofapo Esynthesisandsecretion asdescribed in Methods.The apo Ebandswereexcised from thegels, andradioactivityin apo Eisshown in the lower panel as SEM from triplicate samples.
with [35S ]methionine and quantitatively immunoprecipipating newly synthesized apo E from cell lysates and media. The re-sultsofarepresentative experimentare shown in Table IV. In
culture media, apo E radioactivity was increased more than fourfold in response to a 72-h incubation in TNFa. In cell lysates, there was a 1.8-fold increase in apo Eradioactivity in response to TNFa.
TNFa induction of apo E messenger RNA abundance is associated withphenotypicmodulation. Three markersfor phe-notypic modulationwere assessedto compare withchanges in apoE mRNAexpression duringourexperimental incubations. TNFa,addedat5ng/ml for 24 h, significantly depressedDNA
synthesis (14.5±1.7vs.9.0±0.7 cpmX 103/mgprotein,control
versusTNFa) inTHP1cells. InTableV,expressionof
ICAM-1mRNAispresentedas afunction ofTNFa treatmentduration. This gene codes for an adhesion molecule that is a
counter-receptor for LFA-1 expressed onother leukocytes;interaction between thesetwoproteinsmay enhanceleukocyterecruitment andcommunication(23). As shown, after 6 h in TNFa, ICAM-1 messenger RNA species remained undetectable. However, after 18 and 24h,therewas anincrease in ICAM-1 messenger RNAabundanceproduced byTNFatreatment. TNFa treatment
is also associated witharapidandtransient enhancement of
c-fos geneexpression (TableV). Levels forc-fosmessagewere
undetectable in untreated cells. However, these becameeasily detectable after 15 min in TNFa and remained elevated for
at least 30 min. Levels of the c-fos message again became undetectable after 60 min ofTNFa treatment.
Transduction pathwaysforTNFa
effect
onapo E geneex-pression. Many of the effects ofTNFa have been ascribedto asecond messengerpathwaythatusesceramideasan
important
regulatory molecule (24-26).Theseeffectscanbereproduced by treating cells with cell permeable analogues of ceramide, andwe used this approach to evaluate the importance of this second messenger for transducing the effect ofTNFa on apo E gene expression. As shown in Fig. 3 (upperpanel), C2-ceramide didnotreproduce the effect ofTNFa on apoEgene expression, nor did it potentiate the effect ofTNFa on this gene. Similarly, C2-ceramide did not reproduce the effect of
TNFa onICAM-1expression.Inthis case, however, theTNFa
effectwaspotentiated by ceramide (Fig. 3, middle panel). Nei-therextending the incubation in C2-ceramide for up to 72 h, addition ofC6- ceramideanalogues,nortreatmentof cellswith
Table V. Time Course Effect of TNFa on c-fos and ICAM-I mRNA Expression
c-fos//l-actin
Control N.D.
TNFa 15 min 0.13
TNFa30 min 0.15
TNFa60 min N.D.
ICAM-l/,6i-actin
Control N.D.
TNFa 6 h N.D.
TNFa 18h 0.41
TNFa 24 h 0.48
20 ug of total cellular RNA was isolated from untreated cells or cells treated with5ng/mlofTNFafortheindicated times and analyzed by Northern blot hybridization. The intensity of the bands for ICAM-1 or c-fos alongwith,B-actinwasdeterminedby scanning densitometry.N.D. indicates that there was no c-fos or ICAM-1 signal detected for that experimentalcondition. Values shown are the average of duplicate sam-ples, which varied by < 10%.
sphingomyelinasetogenerate endogenous cellular ceramide led
to induction of apoE mRNAabundance (data not shown). TNFa effects can also be mediated by activation of pkC in transduction pathways separate from those involving ceramide (24-26). We therefore examined whether pathway inhibition of pkC activationduringTNFatreatmentwould attentuate apo E gene response. IntheexperimentshQwnin thebottom panel ofFig. 3, treatment withTNFaproducedasignificant increase in apo E gene expression (P < 0.02). Concurrent treatment with thepkC inhibitor H7, 1-(5-isoquinolinesulfonyl)-2-meth-ylpiperazine, completely preventedthis induction however. A similar inhibition of TNFa stimulationwasproducedby using staursporine toinhibitpkC activity (not shown).
EffectofTNFa onapo Egene transcription. We next con-sidered the mechanismby which TNFa increases apoEmRNA abundance.Forothermacrophagegenes, TNFa has been shown
toincrease mRNA abundancebystabilization of mRNAspecies without any change in gene transcription rate. To assess the contribution of increased apo E gene transcription,THP1 cells
were transiently transfected with a chimeric construct
con-taining apo E promoter sequences (from -2300 to +24 bp) fused to aluciferase reporter. Aftertransfection, asingle pool of cells was created tonormalize transfection efficiency, and thispoolwasthenaliquotedfor incubations in TNFatreatment
for 24 and 48h. TNFainduced theexpression of this chimeric
construct2.2-foldat 24 h and 2.1-foldat48 h (Fig.4).
Discussion
Our studies demonstrate that treatment with TNFa stimulates apoEgenetranscription,mRNAabundance,and apoE synthe-sisand secretion inmonocyte/macrophagecells ina phenotype-dependentmanner.Such concordant responseatallmajor regu-latory loci is not demonstrated by all macrophage targets of
abun-0*0
C.S0
4L.'
1I."
e*g. a
a
a
I
a
40
0.50
0.40
0.30
0.20
0.10
0.00
01-Wd
1w_-nw-O
@2-esI"
no_
UP0
control H7 TNFa TNFa+H7
dance inmacrophages, andthisincreased mRNA abundance is due to a change in mRNA stability (21). While our studies establish that increasedapoEgenetranscriptioncontributesto
enhanced apo E mRNA abundance after TNFa treatment, a
Figure 3.Transduction pathways forTNFa acti-vationofapo E geneexpression. Cells cultured withRPMI-1640 containing 0.05% free-endo-toxin BSAweretreated with 5ng/ml of TNFa, 5AM ofC2-ceramide, 10
jig/ml
ofH7, or a combination of these for24 h.Northern blots for apoE, ICAM-1, andf3-actin
wereprocessedas describedpreviously.The upperpanelshows the effect ofTNFaandC2-ceramideonapo E mRNAabundance aftercorrecting for,/-actin
signal.ThemiddlepanelshowstheTNFa and C2-ceramide effectonICAM-1 mRNA abun-danceafterthe samecorrection.N.D.indicates thattherewasnodetectedsignal for that experi-mentalcondition. Values shownare the meanof duplicate samples, which varied by< 10%.The lowerpanelshowstheeffect ofH7inhibitionon TNFastimulationofapo E mRNAabundance aftercorrection for/l-actin
abundance.Values shownarethemean±SEMfromtriplicatesam-ples.
potential role for apoE mRNA stabilization is notexcluded, especially forthe large increase ofapoE mRNA levels
mea-sured after prolonged incubation in TNFa (up to a 13-fold increase after 72h). Alternatively,inclusion of additionalapoE
a
*1 0
a
a
N0 so
N.D. ND.
U
0
a I l
4-0
a
-I
-
1-,L O."
"4)
o
0
04
we1
24 h 48 h
m I control TNFa
gene elements may be necessarytomaximize the transcriptional response of chimeric constructs to TNFa.
Othershave studied the effect ofTNFa onapoEexpression withinconsistent results. Menjuetal.investigated the effect of TNFa on THP1 cells after these cellsweretreated for4dwith PMA, 12-0-tetradecarvolyphorbol-13-acetate (12). Additional
treatmentfor 12 h inTNFaproducednoeffectonapoE mRNA
abundanceorprotein synthesis. Zuckerman treated thioglycol-late-elicited mouse peritoneal macrophages with TNFa for a
longerperiodof time (48h)andfoundan - 50%reduction of
apoEproteinsecretion (10). Both of these reportsarein accord with our observations regarding the phenotype-specific effect of TNFa on apo E gene expression. Monocyte/macrophages aremultipotential cells thatcanrespondtotheirhumoral envi-ronment in divergent ways. Phenotype-dependent modulation of apoEexpressionbyTNFa is consistent with thepleiotropic effects ofTNFathat have beenreportedfor othermacrophage targets (22, 27). Suchaphenotype-dependentresponse to cy-tokine stimulation may be related to activation ofendogenous cytokinesynthesis and secretionbythefullydifferentiated and activatedmacrophage. Forexample,PMAstimulation hasbeen
shown to increase TNFa mRNA expression within 3 h and TNFaproteinsynthesiswithin 10 h in human monocytes (28). However, if endogenous synthesis ofTNFa by differentiated macrophages accounts for the lack of apo E induction after
treatmentwith exogenous TNFa, the suppressionof apoE
ex-pression observed by
Zuckerman
etal.(10)andbyus(Table I) duringprolonged incubations inTNFawouldnotbeexpected. Alternatively, phenotype-specific effects ofTNFa may relatetodifferencesincell-matrixinteractions between monocyteand
macrophage phenotypes. Adhesion ofcells toextracellular ma-trix has been shownto modify their response to TNFa (29). Further,in the vesselwall,itislikelythat thepresenceof other cytokines will be involved in modulating monocyte/macro-phageresponse toTNFa. Interferon-y, forexample, has been showntobeanimportantinfluenceontheresponse of the
IGF-1 expression aftertreatmentof macrophages withTNFa(30).
TNFa treatmentproduced cell cycle arrest asindicatedby
Figure 4. TNFapromotes the expression of the apoEgene promoter.Cellstransfectedwith -2300/+24 apoEpGL-2 were cultured with 10
ng/mlofTNFafor theindicated timesand then harvested for measurement of luciferase activity. Allcells werederivedfrom asingle poolof transfectedcells tonormalizefortransfection
ef-ficiencyasdescribed in Methods. Thedataare shown as themean±SDofluciferase activity (relative light units/mg protein X 10-6)from triplicate samples.
amarkedly reducedrateof DNA synthesis. In addition, TNFa increased the expression of c-fos within 15 min of its addition. Rapid and transient induction of c-fos expression serves as a
markerfor macrophage differentiation and/or activation (31). Also, TNcFa enhanced the expression of the ICAM-1 gene.In arterial wall plaques, ICAM-1 expression is found in arterial smooth muscle cells and macrophages (32), and it has
pre-viously been shown that TNFa induces the expression of ICAM-1 in arterial smooth muscle cells (33). The effect of TNFa onapo E geneexpression cannotbe generalized to all cytokines associated with monocyte/macrophage activationor
differentiation. Brand et al. added PMA and interferon--y and observednoeffecton apo E mRNA abundance in THP1 cells (34). THP1 cells have been showntoexpressMCSFreceptors
(35); however, in ourlaboratory, the addition of MCSF, with
orwithout phorbol ester, to THP1 cells also had noeffect on apoE mRNAabundance. ThisagreeswiththereportofBrand
et al., who made a similar observation regarding the lack of MCSF effecton apoE protein synthesis and secretion inthese cells (34).
Treatingcells with TNFa hasreportedly produced sphingo-myelin turnoveraswell asendogenous cellular ceramide (24-26).Investigationofbiologically important signal transduction pathways for TNFa has focused on the role ofceramide as a secondmessengeras well as on apotentialroleforpkC activa-tion. There areexamples ofpkC activation and ceramide pro-ducing opposing effects in target monocyte/macrophage cells
(26). InHL60 cells,TNFa leads toincreased activity of
mito-gen-activated proteinkinase and the accumulation of nuclease factor-kB(NFkB) (36, 37).Both of these effects can be
repro-duced by exogenously added ceramide analogues. In Jurkat cells,TNFadecreases DNAsynthesis andincreasesNFkB ac-cumulation(38).Inthesecells, ceramide inhibitsDNA
synthe-sis but does notproduce NFkB accumulation, although it
po-tentiates the effect of TNFa on NFkB accumulation. These
observations supportarole for ceramideas asecondmessenger molecule mediatingTNFaeffects. Recently, however,the
questioned. For example, Betts et al. directly measured ceramide levels, and they could find no difference produced by TNFa
treatment in HL60 cells or Jurkat cells, even though NFkB activation was measurable in both cell types (39). Ourdata demonstratethat ceramide is not involved inTNFamodulation of the apo E gene. Addition of cell-permeant ceramide ana-logues did not increase apo E messenger RNA abundance nor did it potentiate the effect ofTNFa onapo E. We also observed that treatment of THP1 cells with sphingomyelinase,togenerate endogenous cellular ceramide, didnotaugment apo E mRNA abundance (not shown). Inhibitors of pkC activation, however, attenuated the effect ofTNFa on apoEgeneexpression, sug-gesting the importance of this pathway for mediating TNFa effects onthe apoEgene. With respectto the mechanism for
TNFa induction of apo E gene transcription, it is noteworthy thatTNFa treatment of THP1 cells induces expression of the c-fos gene. The protein product of this gene isacomponent of the AP1 transcription complex, which we have previously shownto be important for enhanced apo E gene transcription inthemonocyte/macrophage (19).
In summary, our results document a significant role for
TNFainmodulatingmonocyte/macrophageapo E gene expres-sion. TNFa stimulates apo E genetranscription, mRNA abun-dance, andprotein synthesis in the monocyte/macrophage in a
phenotype-specificmanner.This effect is observed in monocyte cellsorcells in theearly stage of monocyte/macrophage
differ-entiation.Infully differentiated/activated macrophages, TNFa
has no effectorreduces apo E mRNA abundance. Macrophages invessel wall lesionsarederived from the migration of blood-borne monocytes thatcrosstheendothelium andtransform,over
time, into fullydifferentiatedmacrophages. Phenotypic
mono-cytes as well as monocyte/macrophages in more advanced stages of differentiation have been identified in the vessel wall intima ofhyperlipemic animals (40, 41). TNFa produced in the vessel wall could significantly modulate theabundance of apo E derived from monocytes/macrophages that are newly recruited into the vessel wall andthereby influence thegenesis of the vessel wall lesion.
Acknowledgments
Theauthors thankBeverly Burgeforpreparingthemanuscript. This work wassupported bygrantHL39653from the National
Insti-tutesofHealth and wasperformed duringthe tenureofagrant-in-aid fromthe American HeartAssociationand Parke-Davis to T.Mazzone,
anEstablishedInvestigator ofthe American HeartAssociation.H. Duan wassupportedin partbyaMedicalSchool Grantfrom Merck.
References
1.Mahley,R. W.1988.ApolipoproteinE:cholesterol transportproteinwith
expandingrolein cellbiology.Science(Wash DC). 240:622-630.
2. Newman, T.C.,P. A.Dawson, L. L.Rudel,and D. L. Williams. 1985.
QuantitationofapolipoproteinE mRNA in the liver andperipheral tissues of nonhumanprimates. J.BioL Chem.260:2452-2457.
3. Basu, S.K.,M. S.Brown,Y. K.Ho, R. J.Havel,and J. L. Goldstein. 1981. Mousemacrophagessynthesizeandsecrete aproteinresemblingapolipoproteinE. Proc.Nati.Acad. Sci. USA. 78:7545-7549.
4.Mazzone, T., H.Gump,P.Diller, and G. S. Getz. 1987. Macrophage free cholesterol content regulates apolipoprotein E synthesis. J. BioL Chem. 262:11657-11662.
5.Rosenfeld,M.E., S.Butler,V. A. Ord,B. A.Lipton,C. A. Dyer, L. K. Curtiss, W. Palinski, and J. L. Witztum. 1993. Abundant expression of apoprotein
E by macrophages in human and rabbit artherosclerotic lesions.Arterioscler.
Thromb. 13:1382-1389.
6.Mazzone, T., and C. Reardon. 1994. Expression of heterologous human apolipoprotein E by J774 macrophages enhances cholesterol efflux to HDL3. J. Lipid Res.35:1345-1353.
7. Werb, Z., and J. R. Chin. 1983. Endotoxin suppresses expression of apopro-tein E by mouse macrophages in vitro and in culture. J. Biol. Chem.
258:10642-10648.
8. Auwerx, J. H., S.Deeb, J. D. Brunzell, R. Peng, and A. Chait. 1988. Transcriptional activation of the lipoprotein lipase andapolipoprotein Egenes accompanies differentiation in some human macrophage-like cell lines. Biochem. 27:2651-2655.
9.Basheeruddin, K., C. Rechtoris, and T. Mazzone. 1992. Transcriptional and post-transcriptional control ofapolipoproteinEgeneexpressionindifferentiating
humanmonocytes. J. Biol. Chem. 267:1219-1224.
10. Zuckerman, S. H., G. F. Evans, and L. O'Neal. 1992.Cytokine regulation
of macrophage apoE secretion: opposing effects of GM-CSF andTGF-/3. Arthero-sclerosis.96:203-214.
11.Clinton, S. K., R. Underwood, L. Hayes, M. L. Sherman, D. W. Kufe, and P. Libby. 1992. Macrophage colony-stimulating factor gene expressionin vascular cells and in experimental and human atherosclerosis. Am. J. Pathol.
140:301-316.
12.Menju, M., S. Tajima, and A. Yamamoto. 1989.Expressionof the apolipo-protein E gene in a human macrophage-like cell line, THP-1. J. Biochem. 106:505-510.
13. Libby, P., and G. K. Hansson. 1991.Involvementof the immune system in human atherogenesis: currentknowledgeandunansweredquestions.Lab. In-vest. 64:5-15.
14.Barath, P., M. C.Fishbein,J.Cao,J. Berenson, R. H.Helfant,andJ. S. Forrester. 1990.Detection and localization of tumor necrosis factor in human
atheroma. Am. J.Cardiol. 65:297.
15.Mazzone, T., and K. Basheeruddin. 1991. Dissociatedregulationof macro-phage LDL receptor and apolipoprotein E gene expression by sterol. J.LipidRes. 32:507-514.
16.Mazzone, T., L.Pustelnikas,and C. A. Reardon. 1992. Post-translational
regulationofmacrophage apoprotein Eproduction. J. Biol. Chem. 267:1081-1087.
17.Chait, A., and T. Mazzone. 1982. Asecretoryproduct of human monocyte-derived macrophagesstimulates low density lipoprotein receptoractivityin arte-rialsmooth muscle cells and skin fibroblasts. Arteriosclerosis. 2:134-141.
18. Lowry,0. H., N. J. Rosebrough, A. L. Farr, and R. J. Randall. 1951. Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193:265-275. 19. Basheeruddin, K., C.Rechtoris,and T. Mazzone. 1993. Evaluation of the role ofApl-like proteins in the enhanced apolipoproteinE genetranscription
accompanying phorbolesterinducedmacrophagedifferentiation. Biochim. Bio-phy. Acta. 1218:235-241.
20.Zhang, Y.,M.Doerfler,T.C.Lee,B.Guillemin,and W. N. Rom.1993.
Mechanisms of stimulation ofinterleukin-1f3 and tumor necrosis factor-a by
mycobacteriumtuberculosis components. J. Clin. Invest. 91:2076-2083. 21. Chicheportiche, Y., and P. Vassalli. 1994. Tumor necrosis factor induces ablock in the cotranslation ofFcyRllbmRNA in mouseperitonealmacrophages. J. Biol.Chem.269:20134-20138.
22. Shieh, J.-H., R. H. Peterson, D. J. Warren, and M. A. Moore. 1989. Modulation ofcolony-stimulatingfactor-i receptors on macrophages by tumor necrosisfactor. J. Immunol. 143:2534-2539.
23.Springer,T. H.1990. Adhesion receptors of the immuno system. Nature
(Lond.). 346:425-434.
24.Kolesnick, R.,and D. W. Golde. 1994. Thesphingomyelinpathway in
tumornecrosis factor andinterleukin-1 signaling. Cell.77:325-328.
25. Heller, R. A., and M. Kronke. 1994. Tumor necrosis factor receptor-mediatedsignalingpathways. J. Cell Biol. 126:5-9.
26.Hannun,Y. A.1994. Thesphingomyelincycleandthe second messenger functionof ceramide.J. Biol. Chem.269:3125-3128.
27.Witsell,A.L., and L. B. Schook. 1992. Tumor necrosis factor a is an autocrinegrowth regulatorduring macrophage differentiation. Proc. Natl. Acad. Sci. USA.89:4754-4758.
28.Sariban,E., K.Imamura,R. Luebbers, and D. Kufe. 1988. Transcriptional and posttranscriptional regulationof tumor necrosis factor gene expression in humanmonocytes. J. Clin. Invest. 81:1506-1510.
29.Nathan, C., and S. Michael. 1991.Cytokines in context. J. Cell Biol. 113:981-986.
30.Lake,F.R., P. W. Nobel, P. M. Henson, and D. W. H. Riches. 1994. Functionalswitching of macrophage responses to tumor necrosisfactor-a(TNFa)
by interferons:implicationsfor thepleiotropic activities of TNFa. J. Clin. Invest. 93:1661-1669.
31. Muller, R. 1986. Cellular and viral fos genes: structure, regulationof
expressionandbiologicalpropertiesof theirencodedproducts.Biochim.Biophys.
32. Printseva,0.Y., M. M. Pecio, and A. M. Gown. 1992. Various cell types inhumanatheroscleroticlesionsexpress ICAM-1. Am. J. Pathol. 140:889-896. 33. Couffinhal, T., C. Duplaa, L. Labat, J. M. D. Lamaziere,C. Moreau,
0.Printseva, and J. Bonnet. 1993. Tumor necrosis factor-a stimulates ICAM-1 expression in human vascular smooth muscle cells. Arterioscler. Thromb. 13:407-414.
34.Brand, K., N. Mackman, and L. K. Curtiss. 1993.Interferon-yinhibits
macrophage apolipoprotein Eproduction by posttranslational mechanisms. J.
Clin.Invest.91:2031-2039.
35. Datta, R., K. Imamura, S. J. Goldman, A. C. Dianoux, D. W.Kufe,and M. L.Sherman. 1992.Functionalexpression of the macrophage colony-stimulat-ingfactor receptor in human THP-1 monocytic leukemia cells. Blood. 79:904-912.
36. Raines,M.A., R. N.Kolesnick,and D. W.Golde.1993.Sphingomyelinase
andceramide activate mitogen-activated protein kinase in myeloid HL-60 cells. J. Biol. Chem. 268:14572-14575.
37. Yang, Z., M. Costanzo, D. W. Golde, and R. N. Kolesnick. 1993. Tumor necrosis factor activation of thesphingomyelinpathway signals nuclear factor kB translocation in intact HL-60 cells. J.Biol. Chem. 268:20520-20523.
38.Dbaibo, G. S., L. M. Obeid, and Y. A. Hannun. 1993. Tumor necrosis
factor-a (TNF-a) signal transduction through ceramide. J. Biol. Chem. 268:17762-17766.
39. Betts, J. C., A. B. Agranoff, G. J. Nabel, and J. A. Shayman. 1994. Dissociation of endogenous cellularceramidefrom NF-kB activation. J. Biol. Chem. 269:8455-8458.
40.Gerrity,R.G., H. K. Naito, M. Richardson, C. J. Schwartz. 1979. Dietary inducedatherogenesisinswine. Am. J.Pathol.95:775-792.
