Proc. Nadl. Acad. Sci. USAVol. 89, pp. 9049-9053, October 1992Cell Biology

Downregulation of interleukin 8 gene expression in humanfibroblasts: Unique mechanism of transcriptional inhibitionby interferon

(cytokine network/tumor necrosis factor/nuclear run-on assay/neutrophil chemotaxis/inflammation)

IGOR C. OLIVEIRA, PETER J. SCIAVOLINO, TAE H. LEE*, AND JAN VILtEKtDepartment of Microbiology and Kaplan Cancer Center, New York University Medical Center, New York, NY 10016

Communicated by David Sabatini, June 18, 1992 (receivedfor review May 4, 1992)

ABSTRACT The chemotactic cytokine interleukin 8 (IL-8)is produced upon stimulation by various agents in many celltypes, including connective-tissue fibroblasts. Tumor necrosisfactor (TNF) and IL-I are potent inducers of IL-8 expression.Earlier we showed that TNF-induced stimulation of IL-8mRNA accumulation in human FS-4 fibroblasts was inhibitedby interferon (3 (IFN-13) or IFN-y. Here we show that thisinhibition is not specific for TNF, since IFN-,f also reducedIL-8 mRNA accumulation induced by IL-1 or the double-stranded RNA poly (SAC). Treatment with IFN-.8 also decreasedTNF-induced IL-8 protein accumulation. Interestingly, theinhibitory effect was much less pronounced when IFN-(3 wasadded 21 hr before TNF. The inhibitory action of IFN-fl onIL-8 mRNA accumulation was undiminished in the presence ofinhibitors of protein synthesis. Nuclear run-on assays demon-strated that IFN-fi caused a marked inhibition of TNF-inducedIL-8 gene transcription; the transcriptional activation of sev-eral other TNF-induced genes was not inhibited by IFN-(3. Theresults suggest that the specific inhibition of the transcriptionalactivation of IL-8 by IFN is due either to a transient inactiva-tion of a factor required for IL-8 transcription or to theactivation of a selective inhibitory factor.

Cytokines exert most of their biological actions by alteringthe level of gene expression in target cells (1). For example,exposure of cells to interferon a/,B (IFN-a/0) leads to rapidtranscriptional activation of a characteristic set of genes (2,3). Other cytokines known to produce the activation of acommon set of cellular genes are tumor necrosis factor (TNF)and interleukin 1 (IL-1) (4, 5). The target genes activated byTNF or IL-1 partly overlap the genes activated by IFN-a/,8or IFN-y (6-8). Much less is known about the inhibitoryactions of cytokines on gene expression. Expression ofc-myc(9, 10), c-fos (11), immunoglobulin heavy-chain ,u mRNA(12), lamins A and C genes (13), and the mitochondrial genesencoding cytochrome b, subunit I of cytochrome c oxidase,and NADH dehydrogenase subunit 5 (14) was inhibited bytreatment with IFN-a/13 or IFN-'y in some cells.An earlier study described a set of genes activated by TNF

in human fibroblasts (8). Some of the genes activated by TNFwere also inducible by IFN-,3 or IFN-y. In contrast, IFN-,Bor IFN-y inhibited the TNF-induced expression of one ofthese genes, termed TSG-1. Sequencing of the TSG-1 cDNArevealed that this gene was identical to the gene encoding animportant inflammatory cytokine known as IL-8 or NAP-1(8). IL-8 is a member ofthe family of 8- to 10-kDa chemotacticcytokines that includes platelet factor 4, IP-10, Gro, macro-phage inflammatory peptides 1 and 2, macrophage chemo-tactic and activating factor, and several others (reviewed inrefs. 15 and 16). In the present study we investigated the

inhibitory actions of IFN-13 on TNF-induced IL-8 gene ex-pression in human fibroblasts. Our data indicate that IFNtreatment leads to a selective depletion of IL-8 mRNA; thisinhibitory effect correlates with a decrease in the rate of IL-8mRNA transcription. The mechanism responsible for thisaction appears to be unique in that the inhibitory effect ofIFNis established very rapidly and it is not diminished in thepresence of inhibitors of protein synthesis, consistent withthe idea that this effect is due to a transient modification ofa preexisting factor.

MATERIALS AND METHODSMaterials. Recombinant human TNF-a (specific activity, 3

x 107 units/mg) and IFN-y (5 x 106 units/mg) were from M.Tsujimoto, Suntory Institute for Biomedical Research, Osa-ka. Escherichia coli-derived recombinant human IFN-f3 (Be-taseron, 2 X 108 units/mg) was from Triton Biosciences,Alameda, CA, and recombinant human IFN-p derived fromCHO cells (Betaferon, 3 x 108 units/mg) was provided byWieland Wolf, Bioferon, Laupheim, Germany. The E. coli-derived IFN-/3 preparation was used unless specified. Re-combinant human IL-la (3 x 108 units/mg) was a gift fromPeter Lomedico and Alvin Stem, Hoffmann La-Roche, Nut-ley, NJ.

Cell Culture. Human diploid FS-4 fibroblasts were main-tained in Eagle's minimum essential medium (GIBCO/BRL)supplemented with 5% fetal bovine serum, 6 mM Hepes, 3mM Tricine, and 50 gg of gentamicin per ml. Cultures wereseeded in 175-cm2 flasks (2 x 106 cells per flask), 75-cm2flasks (8.5 x 105 cells per flask), or 175-cm2 dishes (2 x 106cells per dish) and grown to confluency (9-11 days). Thecultures were then switched to medium with 0.25% fetalbovine serum and incubated for an additional 4-6 days. Allagents used were added directly to the medium, without priormedium change.

Nuclear Run-On Analysis. Nuclear run-on assays wereperformed as described (3). Briefly, FS-4 cells were rinsedand scraped off the dish in ice-cold phosphate-bufferedsaline. The nuclei were collected by centrifugation afterdisruption of the cells with a Dounce homogenizer (B pestle)and washed once in reaction buffer (20 mM Tris, pH 7.5/10mM MgCl2/140 mM KCl/20% glycerol/0.5 mM dithiothrei-tol). The nuclei were then resuspended in the same buffercontaining 0.5 mM ATP, CTP, and GTP, 5 ILM UTP, and0.2-0.3 mCi of [a-32P]UTP (3000 Ci/mM; ICN; 1 Ci = 37GBq), and the elongation reaction was allowed to proceed for

Abbreviations: IFN, interferon; IL, interleukin; TNF, tumor necro-sis factor.*Present address: Department of Molecular Biology, MassachusettsGeneral Hospital, Boston, MA 02114.tTo whom reprint requests should be addressed at: Department ofMicrobiology, New York University Medical Center, 550 FirstAvenue, New York, NY 10016.

9049

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertisement"in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Dow

nloa

ded

by g

uest

on

June

22,

202

1

Proc. Natl. Acad. Sci. USA 89 (1992)

15 min at 300C. The templates were digested with 200 units ofRNase-free DNase I (Boehringer Mannheim), and the labeledRNA was extracted with 650C phenol and phenol/chloroformand precipitated with trichloroacetic acid. Plasmids contain-ing the sequences of interest were alkali-denatured andimmobilized on a nylon membrane (Schleicher & Schuell).After prehybridization for 2-3 hr in TESS2D (10mM Tes, pH7.4/2x Denhardt's solution/400 mM NaCl/4 mM EDTA/0.1% SDS with 200 1Lg ofE. colitRNA per ml), the filters werehybridized overnight with the labeled nuclear RNA probe.Northern Blot Analysis. Total cytoplasmic RNA was iso-

lated (8), and samples (10 ,ug) were subjected to electropho-resis in a formaldehyde/Mops/1% agarose gel and trans-ferred to a nylon membrane (Schleicher & Schuell). Blotting,prehybridization, hybridization, and washing conditionswere as described (17). The TSG-1/IL-8 cDNA probe (8) andthe pHe7 cDNA probe (from P. B. Sehgal, RockefellerUniversity, NY) were 32P-labeled with a nick-translation kit(Boehringer Mannheim). The filters were exposed to a KodakXAR-5 film with two intensifying screens at -700C. Furtherquantification was performed with the aid of a laser densi-tometer. The results are expressed as relative mRNA accu-mulation, with pHe7 mRNA as the internal standard.

RESULTSInhibitory Effect of IFN-P and IFN-y on IL-8 mRNA

Accumulation Induced by TNF or Other Agents. Earlier wereported that the induction of IL-8 mRNA by TNF in humanFS-4 fibroblasts was inhibited when cells were treated simul-taneously with TNF and IFN-P or IFN-y (8). This inhibitionwas highly specific for IL-8, since IFN-P or IFN-y showed noinhibitory action on the ability ofTNF to induce seven othertypes of mRNA in the same cells. We examined the effectsof different doses of IFN-13 or IFN-'y on IL-8 mRNA levelsin FS-4 cells exposed to TNF and IFN-/3 or IFN-'y for 2 hr(Fig. 1). IFN-,3 or IFN-y produced a dose-dependent de-crease in the steady-state levels of TNF-induced IL-8mRNA, with IFN-f3 showing a more potent inhibitor effect atlower doses than IFN-y. An inhibitory effect of -90o wasproduced with IFN-P at -'100 units/ml or IFN-y at -500units/ml. However, the maximal inhibition of IL-8 mRNAlevels produced by IFN treatment varied somewhat from

100,1

r 80-0

E' 60*0

z

E 40

t0C20

h ~

8 32 128IFN, units/mi

FIG. 1. Inhibitory effects of IFN-,3 and IFN-y are dose-dependent. FS-4 cells were treated with TNF-a (20 ng/ml) andvarious doses of IFN-(3 (o) or IFN-y (.) for 2 hr. Total RNA wasextracted, and samples were subjected to Northern blot analysis andquantitation by laser densitometry.

experiment to experiment. Only IFN-pB was used in subse-quent experiments.To determine whether the inhibitory action ofIFN-/3 on IL-8

mRNA accumulation was specific for induction by TNF, weexamined the action of IFN-,B on IL-8 mRNA levels inducedby IL-la or poly(IC) (8). FS-4 cells were exposed for 2 hr toTNF, poly(I-C), or IL-la in the presence or absence of IFN-,3(Fig. 2). The inhibitory effect of IFN-13 on IL-8 mRNAaccumulation was pronounced with all inducing agents usedand, hence, is not specific for TNF. It is noteworthy thatpoly(IC) is best known as a potent inducer of IFN-,8 (18, 19),but since IFN-p inhibits rather than stimulates IL-8 induction,the ability of poly(I-C) to induce IL-8 mRNA must be separatefrom its IFN-(3-inducing capacity.

Inhibitory Effect of IN-1 on TNF-induced IL-8 ProteinAccumulation. Cultures were treated with TNF alone orsimultaneously with TNF and IFN-f3. Culture fluids werecollected at intervals for the determination of IL-8 protein bya specific immunoassay (Fig. 3). IL-8 was first demonstratedin the culture supernatants 3-6 hr after stimulation with TNFand continued to increase throughout the duration of theexperiment (% hr). The addition of IFN-(3 (100 units/ml)reduced the levels ofIL-8 protein in the culture medium at thevarious time points examined by 60-75%.

Transient Nature of the Inhibitory Effect of IFN-P on IL-8mRNA Accumulation. Most actions of IFN are due to astimulation of expression of selected genes in the target cells(20-22). If the inhibitory action of IFN-/3 on IL-8 mRNAaccumulation also were mediated by IFN-induced protein(s),one would expect that preincubation of cells with IFN-f,followed by the addition of TNF, would lead to a greaterinhibition of IL-8 mRNA accumulation than simultaneoustreatment of cells with IFN-p and TNF for 2 hr. To elucidatethis point, cells were first treated with IFN-f3 for 1 or 4 hr.Thereafter, TNF was added and the cultures were incubatedin the presence of both IFN-,B and TNF for 2 hr, at whichpoint steady-state IL-8 mRNA levels were determined (Fig.4). In parallel, one group of cells was treated simultaneouslywith TNF and IFN-,f for only 2 hr. Surprisingly, the inhib-itory effect of IFN-,B decreased with the length of preincu-bation, and little inhibition was seen when IFN-P was added4 hr before TNF.One possible explanation of these results was that IFN-3

somehow inhibited the generation of a signal required forIL-8 induction by TNF, but only when IFN-/ and TNF wereadded at about the same time. To test this idea we firstexposed FS-4 cells to TNF for 2 or 4 hr and then addedIFN-f3. IL-8 mRNA levels were determined by Northern blotanalysis 2 or 4 hr after the addition of IFN-13. Compared withcells exposed to TNF alone, there was a marked decrease inIL-8 mRNA in the IFN-,3-treated cultures (data not shown),indicating that the inhibitory action ofIFN-P was not directedselectively at an early event in the stimulation of IL-8 geneexpression.

Reduction in IL-8 mRNA by lFN-13 Is Undiminished in thePresence of Inhibitors of Protein Synthesis. Earlier we showedthat the induction of IL-8 mRNA by TNF was direct, becausecycloheximide not only failed to reduce but in fact "super-induced" IL-8 mRNA levels (8). A similar superinductionwas seen in the experiment shown in Fig. 5. Interestingly, theinhibitory effect of IFN-,8 on TNF-induced IL-8 mRNA alsowas not reduced in the presence ofcycloheximide, suggestingthat this action too did not require protein synthesis. Inaddition, anisomycin, a protein synthesis inhibitor acting bya mechanism different from that of cycloheximide, alsocaused a superinduction of IL-8 mRNA by TNF and failed todiminish the inhibition of IL-8 mRNA induction by IFN-f3(data not shown).IFN-P Inhibits the Rate of IL-8 mRNA Transcription.

Nuclear run-on assays were carried out in FS-4 fibroblasts

9050 Cell Biology: Oliveira et al.

r--

Dow

nloa

ded

by g

uest

on

June

22,

202

1

Proc. Natl. Acad. Sci. USA 89 (1992) 9051

A100 -

80 -

60 -

40 -

*,^.D

ZL ILZ

IL

zI~-

20 -

0 -

0b-O-

cJ0(.)

0

0.

B

- goz Z_L _L

FIG. 2. IFN-,B can reduce IL-8 mRNA accumulation elicited by various inducers. RNA was extracted from all groups 2 hr after the initiationof treatment and analyzed by Northern blotting followed by laser densitometry. (A) FS-4 cells were treated for 2 hr with TNF-a (20 ng/ml),or poly(I1C) (50 Aug/ml) in the presence or absence of IFN-f3 (500 units/ml). (B) FS-4 cells were treated for 2 hr with IL-la (3 ng/ml) in the presenceor absence of IFN-,8 (500 units/ml).

treated for 2 hr with TNF alone, TNF in the presence ofIFN-/3, or IFN-f3 alone (Fig. 6A). No IL-8 transcription wasdetected in untreated or IFN-f3-treated cells. Treatment withTNF resulted in a transcriptional induction of IL-8 message.Cells treated simultaneously with TNF and IFN-13 showed areduced transcriptional activation of IL-8 (about 25% of thatseen in cultures treated with TNF alone, according to den-sitometric determination). To prove that simultaneous treat-ment with TNF and IFN-p did not cause a general inhibitionof transcription, the same nuclear RNA probes were alsohybridized with ,B-actin DNA (provided by E. B. Ziff, NewYork University Medical Center) and with cDNAs specificfor methallothionein II/TSG-37 and TSG-6 (8) and for class

15-

EC 10-0-

5-

I histocompatibility antigen HLA-B7 (23). IFN-f3 did notinhibit the transcription of TNF-induced TSG-6 mRNA, andtranscription of the metallothionein II and HLA-B7 genes,known to be inducible by IFN-,B (21, 22), was in factenhanced by IFN-,8. Thus simultaneous addition ofTNF and

100

c

0° 80-a

EO 60

a

zE 40-E

a 20i

w

-

0

CM CM Cd4 C4

IL z U. L.Z IL Z ZI.- I.- I.-

IL A A

z

-r

z zZL ZL

Ficlationthe picultuldeterMinn

0 20 40 60 80 100 FIG. 4. Inhibitory effect of IFN-P on TNF-induced IL-8 mRNATime, hr accumulation: effect of IFN-p added before the onset of treatment

with TNF. FS-4 cells were left untreated or treated with TNF (203. 3. Effect of IFN-,8 on TNF-induced IL-8 protein accumu- ng/ml), IFN-/3 (100 units/ml), or both TNF and IFN-,8 for the timesi. FS-4 cells were treated with TNF-a (20 ng/ml) alone (e) or in indicated. Arrow indicates that IFN-,3 was added 1 or 4 hr beforeresence of IFN-j3 (100 units/ml) (o). At the indicated intervals, TNF; in the latter groups TNF was added to the IFN-p3-containingre fluids were collected and the IL-8 protein concentration was medium, and further incubation for 2 hr was in the presence of bothmined by immunoassay (Quantikine ELISA kit, R&D Systems, IFN-,B and TNF. RNA was extracted and analyzed by Northernieapolis). blotting and laser densitometry.

100

0_ 80'U

Eo 60Ua

Z 40E

20

0m0

Cell Biology: Oliveira et al.

Dow

nloa

ded

by g

uest

on

June

22,

202

1

Proc. Natl. Acad. Sci. USA 89 (1992)

100 -

0co

EUuuco

z

mES

coi

80 -

60 -

40 -

20 -

0 -0b-

cJ0

A

C

L. X 8

a--ctin _

p; T Z

MT-11

TSG-6-

HLA-B7

L z xZ ILL Z

+ +

IL ,ILU-. ZLI.-

zIL.

x

Lz

B

o

o z

FIG. 5. IFN-,3 inhibits TNF-induced IL-8 mRNA accumulationin the presence of cycloheximide. FS-4 cells were treated for 2 hrwith TNF (20 ng/ml) alone or in the presence of IFN-,B (100units/ml), in either the presence or the absence of cycloheximide(CHX) (10 Ag/ml). Thereafter, RNA was isolated and analyzed byNorthern blotting followed by laser densitometry.

IFN-,B did not produce a general transcriptional downregu-lation.The inhibitory effect of IFN-,3 on IL-8 transcription was

maximal when IFN-13 and TNF were added simultaneously(Fig. 6B). Addition of IFN-f3 2 hr before TNF diminished theinhibitory effect, and its addition 6 hr earlier abolished theeffect. These results are in general agreement with the datashown in Fig. 4, further illustrating that when IFN-/3 is addedto cells prior to TNF, the inhibitory effect is relativelyshort-lasting.

DISCUSSIONOne characteristic feature of many cytokines is that theiractions involve intricate interacting networks or cascades. Inmany instances the interactions are stimulatory; i.e., cyto-kines stimulate the production of other cytokines in the targetcells (8, 24). Less frequently, a cytokine network involves aninhibitory action on cytokine production (1). One such inter-action, the inhibition of IL-8 production by IFN (8), has beenanalyzed in the present study. The decrease in steady-statemRNA levels seen after IFN-p treatment was matched by adecrease in the IL-8 mRNA signal generated in the run-onassay, indicating an inhibition at the level of transcription.Most IFN actions are mediated by intracellular proteins

induced by IFN in the target cells (20-22). It seemed plausiblethat the inhibitory action on IL-8 transcription also could bemediated by IFN-induced protein(s). To begin to address thequestion whether an IFN-induced protein(s) is involved, weexamined the action of IFN-P3 added at different times beforeTNF. The inhibition was the greatest when IFN-3 and TNFwere added at the same time, and the effect on steady-stateIL-8 mRNA levels (Fig. 4) or on IL-8 transcription (Fig. 6B)decreased with the increase in the time between IFN-,/addition and TNF addition. The fact that the inhibition wasmarked when IFN-/3 and TNF were both present for only 2

FIG. 6. Effect of IFN-,/ on TNF-induced IL-8 transcriptiondetermined by nuclear run-on analysis. (A) FS-4 cells were either leftuntreated or treated for 2 hr with TNF (20 ng/ml), CHO cell-derivedrecombinant human IFN-,f (500 units/ml), or both TNF and IFN.MT, metallothionein. (B) Agents used were the same as for A and thetimes oftreatment were as indicated. Arrow indicates that IFN-,8 wasadded 2 or 6 hr before TNF; in the latter groups TNF was added toIFN-p3-containing medium, and further incubation was in the pres-ence of both IFN-P3 and TNF. GAPDH, glyceraldehyde-3-phosphatedehydrogenase.

hr, together with the rapidly diminishing effectiveness ofIFN-f3 when it was added to cells before TNF, speaks againstthe involvement of IFN-j-induced protein(s). In addition, theinhibitory action of IFN-/3 on TNF-induced IL-8 mRNAlevels was not reduced in the presence of inhibitors of proteinsynthesis, cycloheximide (Fig. 5) and anisomycin (data notshown). If the inhibitory action on IL-8 transcription is notmediated by de novo synthesized IFN-induced protein(s), itcould be due to the modification of preexisting factor(s). Themodification brought about by IFN might lead to the gener-ation of factor(s) that act as specific transcriptional inhibitorsof the IL-8 gene. Mukaida et al. (25), who studied cis-actingelements responsible for transcriptional activation of IL-8 byTNF, IL-1, or phorbol 12-myristate 13-acetate in a humanfibrosarcoma cell line, concluded that two elements arenecessary and sufficient for IL-8 induction: an NF-KB-likeand an NF-IL-6-like element. Whether IFN might affect theprotein factors recognized by these elements remains to beexamined.Although the data in Fig. 6 indicate that IFN-f3 inhibits

TNF-induced transcription of IL-8 mRNA, these results donot rule out the possibility that IFN-3 affects IL-8 synthesisalso at some other point(s). Despite the transient nature of theinhibitory action of IFN-pS, accumulation of IL-8 proteinremained significantly reduced in the presence of IFN-/3 over

IL- 8

pG EM

GAPDH

9052 Cell Biology: Oliveira et al.

LL

Dow

nloa

ded

by g

uest

on

June

22,

202

1

Proc. Natl. Acad. Sci. USA 89 (1992) 9053

a period of at least 4 days (Fig. 3). One possible explanationis that in addition to inhibiting IL-8 mRNA transcription,IFN-,f also blocks IL-8 synthesis at a posttranscriptional,translational, or posttranslational level. It is not unusual fora single agent to inhibit the synthesis of a protein at more thanone level (26). That IFN-,3 also might destabilize IL-8 mRNA,in addition to inhibiting its transcription, was suggested by arapid decrease in the IL-8 mRNA level seen when IFN-f3 wasadded to FS-4 cells 4 hr after TNF, despite the fact that inactinomycin D-treated cells IL-8 mRNA was stable forseveral hours (data not shown).Although IFN-a/,8 and IFN-y are best known as activators

of gene expression (20-22), the IL-8 gene is not the only genewhose expression is inhibited by IFN. One gene on whichIFN was shown to exert an inhibitory effect is c-myc inhuman lymphoblastoid (Daudi) cells (9, 10). Whether thiseffect was due to an inhibition of transcription (10) or to adecrease in the stability of c-myc mRNA (9, 27) has not beenfully resolved. It does not appear that the inhibition of c-mycexpression and the inhibition of IL-8 gene expression aremediated by the same IFN-activated mechanism. The inhib-itory action ofIFN-a or IFN-,B on c-myc was seen in cells thatwere exposed to IFN for 20-48 hr (9, 10). In contrast, theinhibitory effect on IL-8 mRNA diminished rapidly with thelength of incubation of cells with IFN-f3 (Figs. 4 and 6B).Other genes whose expression was shown to be inhibited byIFN include the c-fgr (28) and c-fos (11) protooncogenes,immunoglobulin heavy-chain Au mRNA (12), lamins A and Cgenes (13), and a set of mitochondrial genes (14). Most ofthese systems differ from the one described here. Thus, theinhibitory effect on c-fos mRNA seen with IFN-y in murinemacrophages, was found to be due to a decrease in mRNAstability and not to an effect on transcription (11). Theinhibitory action on heavy-chain ,u mRNA, seen with IFN-ain the Daudi lymphoblastoid cell line, was thought to occurat the level of termination and/or posttranscriptional pro-cessing of mRNA (12). The inhibitory action of IFN-a/8 onthe expression of several mitochondrial genes is differentfrom our system in the unique nature of the target genes andin that the former inhibitory effect is completely blocked inthe presence of cycloheximide (14). The mechanisms of theinhibitory action of IFN on the expression of some othergenes (13, 28) have not been analyzed, making a comparisonwith our data difficult.We believe that the inhibitory effect on IL-8 synthesis

represents a heretofore unknown mechanism of IFN actionbecause (i) the inhibitory effect is on the induced, notconstitutive, expression of a gene; (ii) the inhibition ispronounced at the level of transcription; (iii) establishment ofthe inhibitory action is unusually rapid and is not diminishedin the presence of inhibitors of protein synthesis; and (iv) themodification brought about by IFN is short-lasting. In addi-tion to the FS-4 line of human diploid foreskin fibroblasts, wedemonstrated an inhibition of steady-state accumulation ofTNF-induced IL-8 mRNA by IFN-,B in the WI-38 line ofhuman diploid lung embryo fibroblasts (data not shown). Aninhibitory effect of IFN-y on IL-8 production in humanthymic epithelial cells was also reported (29). However,several virus-transformed or tumor-derived cell lines, inwhich IL-8 mRNA levels were induced by TNF, provedrefractory to the inhibitory action of IFN-f8 on IL-8 mRNA(data not shown). Hence, it appears that the mechanismresponsible for the inhibition of IL-8 mRNA synthesis is notoperative in most transformed cells. The most importantfunction of IL-8 is thought to be its ability to recruit neutro-phils and T lymphocytes to inflammatory sites (15, 16). BothIFN-ci/3 and IFN-y can be produced at inflammatory sites

inflammatory activity in animal models of delayed-type hy-persensitivity (20). These findings suggest that the inhibitoryaction of IFN-a/f3 and IFN-y on IL-8 synthesis may havepathophysiological significance in inflammation.

We thank Drs. Richard Pine, Kouji Matsushima, Naofumi Mu-kaida, and Jedd Wolchok for advice; Dr. Sherman Weissman for theHLA-B7 probe; Angel Feliciano for technical assistance; and IleneTotillo for preparation of the manuscript. This work was supportedby grants from the National Institutes of Health (R35CA49731) andthe American Cancer Society (CD-477). I.C.O. is supported byfellowship 1335189-3 from the Brazilian Ministry of Education(CAPES), and P.J.S. by a predoctoral fellowship from NationalInstitutes of Health Grant T32AI07180.

1. Thomson, A. W. (1991) The Cytokine Handbook (Academic,London).

2. Friedman, R. L., Manly, S. P., McMahon, M., Kerr, I. M. &Stark, G. R. (1984) Cell 38, 745-755.

3. Larner, A. C., Jonak, G., Cheng, Y.-S. E., Korant, B., Knight,E. & Darnell, J. E., Jr. (1984) Proc. Natl. Acad. Sci. USA 81,6733-6737.

4. Vildek, J. & Lee, T. H. (1991) J. Biol. Chem. 266, 7313-7316.5. Dinarello, C. (1991) Blood 77, 1627-1652.6. Beresini, M. H., Lempert, M. J. & Epstein, L. B. (1988) J.

Immunol. 140, 485-493.7. Rubin, B. Y., Anderson, S. L., Lunn, R. M., Richardson,

N. K., Hellerman, G. R., Smith, L. J. & Old, L. J. (1988) J.Immunol. 141, 1180-1184.

8. Lee, T. H., Lee, G. W., Ziff, E. B. & Vildek, J. (1990) Mol.Cell. Biol. 10, 1982-1988.

9. Knight, E., Jr., Anton, E. D., Fahey, D., Friedland, B. K. &Jonak, G. J. (1985) Proc. Natl. Acad. Sci. USA 82, 1151-1154.

10. Einat, M., Resnitzky, D. & Kimchi, A. (1985) Nature (London)313, 597-600.

11. Radzioch, D. & Varesio, L. (1991) Mol. Cell. Biol. 11, 2718-2722.

12. Meurs, E. & Hovanessian, A. G. (1988) EMBO J. 7,1689-1696.13. Alldridge, L. C.,O'Farrell, M. K.&Dealtry,G. B. (1991)Exp.

Cell Res. 195, 546-550.14. Shan, B., Vazquez, E. & Lewis, J. A. (1990) EMBO J. 9,

4307-4314.15. Oppenheim, J. J., Zachariae, C. 0. C., Mukaida, N. & Mat-

sushima, K. (1991) Annu. Rev. Immunol. 9, 617-648.16. Van Damme, J. (1991) in The Cytokine Handbook, ed. Thom-

son, A. W. (Academic, London), pp. 201-214.17. Lin, J.-X. & Vildek, J. (1987) J. Biol. Chem. 262, 11908-11911.18. Field, A. K., Tytell, A. A., Lampson, G. P. & Hilleman, M. R.

(1967) Proc. Natl. Acad. Sci. USA 58, 1004-1010.19. Havell, E. A., Berman, B., Ogburn, C., Berg, K., Paucker, K.

& Vildek, J. (1975) Proc. NatI. Acad. Sci. USA 72, 2185-2189.20. De Maeyer, E. & De Maeyer-Guignard, J. (1988) Interferons

and Other Regulatory Cytokines (Wiley, New York).21. Vildek, J. (1990) in Handbook of Experimental Pharmacology,

eds. Sporn, M. B. & Roberts, A. B. (Springer, Berlin), Vol.95/11, pp. 3-38.

22. Sen, G. C. & Lengyel, P. (1992) J. Biol. Chem. 267,5017-5020.23. Collins, T., Lapierre, L. A., Fiers, W., Strominger, J. L. &

Pober, J. S. (1986) Proc. Natl. Acad. Sci. USA 83, 446-450.24. Van Damme, J., Decock, B., Conings, R., Lenaerts, J.-P.,

Opdenakker, G. & Billiau, A. (1989) Eur. J. Immunol. 19,1189-1194.

25. Mukaida, N., Mahe, Y. & Matsushima, K. (1990) J. Biol.Chem. 265, 21128-21133.

26. Beutler, B., Krochin, N., Milsark, I. W., Luedke, C. & Ce-rami, A. (1986) Science 232, 977-980.

27. Dani, C., Mechti, N., Piechaczyk, M., Lebleu, B., Jeanteur, P.& Blanchard, J. M. (1985) Proc. Natl. Acad. Sci. USA 82,4896-4899.

28. Sharp, N. A., Luscombe, M. J. & Clemens, M. J. (1989) On-cogene 4, 1043-1046.

29. Galy, A. H. M. & Spits, H. (1991) J. Immunol. 147, 3823-3830.30. Heremans, H., Dijkmans, R., Sobis, H., Vandekerckhove, F.

(20, 30). Moreover, under some conditions IFNs have anti-

Cell Biology: Oliveira et al.

& Billiau, A. (1987) J. Immunol. 138, 4175-4179.

Dow

nloa

ded

by g

uest

on

June

22,

202

1