Multiple biological activities of human recombinant interleukin 1

Multiple biological activities of human

recombinant interleukin 1.

C A Dinarello, … , P E Auron, R C Reuben

J Clin Invest.

1986;

77(6)

:1734-1739.

https://doi.org/10.1172/JCI112495

.

Complementary DNA coding for human monocyte interleukin 1 (IL-1), pI 7 form, was

expressed in Escherichia coli. During purification, IL-1 activity on murine T cells was

associated with the recombinant protein. Homogeneous human recombinant IL-1 (hrIL-1)

was tested in several assays to demonstrate the immunological and inflammatory properties

attributed to this molecule. hrIL-1 induced proliferative responses in a cloned murine T cell

in the presence of suboptimal concentrations of mitogen, whereas no effect was observed

with hrIL-1 alone. At concentrations of 0.05 ng/ml, hrIL-1 doubled the response to mitogen (5

X 10(6) half maximal units/mg). Human peripheral blood T cells depleted of adherent cells

underwent a blastogenic response and released interleukin 2 in the presence of hrIL-1 and

mitogen. hrIL-1 was a potent inflammatory agent by its ability to induce human dermal

fibroblast prostaglandin E2 production in vitro and to produce monophasic (endogenous

pyrogen) fever when injected into rabbits or endotoxin-resistant mice. These studies

establish that the dominant pI 7 form of recombinant human IL-1 possesses immunological

and inflammatory properties and acts on the central nervous system to produce fever.

Research Article

Multiple Biological Activities of Human Recombinant Interleukin

I

Charles A. Dinarello,* Joseph G.Cannon,* JamesW.Mier,*Harry A.Bernheim,4 GailLoPreste,* Debra L.Lynn,§

Richard N. Love,§AndrewC.Webb,"1Philip E. Auron,*¶ Roberta C. Reuben,** Alexander Rich,¶

Sheldon M. Wolff,*andScottD. Putney§

*Department ofMedicine, TuftsUniversitySchool ofMedicine and theNewEngland MedicalCenter, Boston,Massachusetts0211 ,;

tDepartmentofBiology, TuftsUniversity,Medford, Massachusetts02155;§Repligen Corporation, Cambridge,Massachusetts 02139; 'DepartmentBiological Sciences, Wellesley College, Wellesley,Massachusetts 02181; ¶DepartmentofBiology, MassachusettsInstituteof

Technology, Cambridge, Massachusetts02139; **Cistron Technology, Inc., PineBrook,New Jersey07053

Abstract

Complementary DNA coding for human monocyte interleukin1

(IL-1), pl 7 form, was expressed in Escherichia coli. During purification,IL-1 activity on murine T cells wasassociated with the recombinant protein. Homogeneous humanrecombinant

IL-1(hrIL-1) was tested in several assays to demonstrate the im-munological and inflammatory properties attributed to this mol-ecule. hrIL-1induced proliferative responses in a cloned murine Tcellin the presence of suboptimal concentrations ofmitogen, whereas no effect was observed with hrIL-1 alone. At concen-trationsof 0.05ng/ml,hrIL-1 doubled the response tomitogen (5 X 106 half maximal units/mg). Human peripheral blood T cellsdepletedofadherentcellsunderwent ablastogenic response and releasedinterleukin 2 in the presence ofhrIL-1 and mitogen. hrIL-1 was a potentinflammatory agent by its ability to induce human dermal fibroblast prostaglandin E2 production in vitro and toproduce monophasic (endogenous pyrogen) fever when injectedintorabbitsorendotoxin-resistant mice. These studies establishthatthedominantpl7form of recombinant human IL-1 possesses

immunological

and

inflammatory

propertiesandacts onthe central nervous system to produce fever.

Introduction

Although the activatedmonocyte/macrophageproduces a great number and variety of biologicallyactive molecules (1), consid-erableattention hasfocusedon interleukin I

(IL-1)'

as an im-portant mediator ofhost responses to microbial and inflam-matory diseases. There presently is increasing evidence that

IL-1 is theprimary inducerof many components of systemic

"acute-phase"

responses andis also responsible forlocal inflam-matorychangesinvarious jointdiseases (2-4). There are many biological activities attributedtoIL-1 includingits effects on the endocrinologic, nervous, andimmunological systems. In

addi-Address reprint requests to Dr. Dinarello, Tufts University School of

Medicine, 136 HarrisonAvenue,Boston, MA 021 1 1.

Receivedfor

publication

15November 1985 and in revisedform 10 February 1986.

1.Abbreviationsused in thispaper:CTLL, cytotoxic Tlymphocyteline; FCS,fetalcalfserum;hrIL-1,humanrecombinant interleukin1; IL-1,

interleukin 1; LAF, lymphocyteactivating factor; NA, nonadherent; PGE2, prostaglandinE2;PHA,phytohemagglutinin;SDS-PAGE, sodium dodecylsulfate-polyacrylamide gel electrophoresis.

tion, in mesenchymal tissue such as cartilage, bone, muscle, and fibrous tissue, IL-I induces changes associated with both de-structiveaswellasreparativefunctions. Howasinglemolecule canaffect these seemingly opposite activities in a wide variety ofcells remainsachallenging biological question.

Initial experimentssuggestingabroadrole for IL-I in me-diatinghost responsestoinfection andinjurydemonstrated that highly purified monocyte-derived endogenous pyrogens were abletoaugment T cellproliferationto mitogens and antigens,

aproperty knownaslymphocyte activating factor (LAF) (5, 6); inadditiontocausingfever, IL-1 induced severalcomponents ofthe acute phase response inexperimental animals (7). Recent studies from several laboratories, using IL- 1 purified from mononuclear phagocytestoapparenthomogeneityasjudgedby sodiumdodecyl

sulfate-polyacrylamide gel electrophoresis

(SDS-PAGE) haveconfirmedtheinitialobservationsonthe multiple biologicalactivities ofIL- 1 (8-1 1). Multiple biologicalactivities ofIL- I arealsoassociatedwithmolecules

eluting

from gel-fil-trationchromatography at30,000-38,000 molwt(12-14) and withmoleculesfocusingatdistinct isoelectricpoints (- 7, 6,and 5) (6, 10, 12, 13); antibodies developedto individual charged formsdo notcross react(15). In addition, small fragmentsof 4,000 and 2,000 molwtIL-1 derived from humansources have

been described and raise thepossibilitythat the activesite for some biological activities is segregated in small peptides (16-18).

The sequence ofacDNAcodingfor human monocyte IL-I has beenrecently described ( 19); the derived polypeptide

se-quence isaprecursor form of31,000molwtandapl near5, whereas the maturecleavagepeptide has a molecular weight of

17,500 and a pl near 7.Although another human monocyte

IL-IcDNAhas been reported(20),it shares nearly 80% homology withtheaminoacidsequencereportedforthemurineIL-1 from the P388D cell line(21), which hasapI of 5. It is clear from

bothproteinand messengerRNA(mRNA)measurements that

the pI7form of human IL- I is thepredominantform. As much

as5%of the total polyAmRNA extracted from human

mono-cytesstimulatedwithendotoxincodes for this form of IL- 1 (22) and this form is also the predominant IL-l found in human pathological fluids.Inthis reportwedescribemultiple biological activities ofpurifiedhuman recombinant IL-l (hrIL-l),which wasexpressed in E. coli, and contained amino acids 71-269 (24,500 mol wt)or 112-269(17,500 molwt)of the precursor polypeptide (31,000molwt).Therecombinant

proteins

exhibit severalbiological propertiesassociated with theimmunological andinflammatory propertiesofthe molecule.

Methods

Materials.Phytohemagglutinin (PHA)was_purchasedfrom

Burroughs-Wellcome &Co.(ResearchTriangle, NC).All endotoxin determinations

J.Clin. Invest.

©TheAmericanSocietyfor ClinicalInvestigation,Inc.

0021-9738/86/06/1734/06 $1.00

were madeusing the Limulus test usinglysatewith a sensitivity of20

pg/mi obtained from Associates ofCapeCod (Woods Hole, MA).

Po-lymyxinB waspurchasedfrom PfizerInc., New York, NY anddissolved inpyrogen-freewater.

Purificationofhumanmonocyte-derived IL-I. Human platelet-pher-esisbyproducts were used as a sourceofmononuclear cells; the mono-nuclear cells wereadjustedto a concentrationof I X 107/mlin RPMI-1640(Microbiological Associates,Inc., Bethesda, MD)without methi-oninecontaining 1%dialyzed humanAB serum.Cellswereincubated

at37°Cfor 1.5 h and the nonadherent cellswereremoved byvigorous shaking. The adherent cellpopulationwasstimulated with opsonized heat-killedStaphylococcus albus in the presence of methionine-free me-dium containing50,Ci/mlof [35S]methionineandincubated for 36h. Thesupernate wasremoved,centrifugedat 3,000g,andfiltered. The supernatewaspurified bysequentialimmunoabsorption,gel-filtration,

andchromatofocusing. Details of the purification procedures and the antibodyused tomaketheimmunoabsorbent havebeenpublished pre-viously(10,23).IL-I withpl 6.8-7.2 elutedfromthechromatofocusing

step ashomogeneous 17,500 molwtdetermined by SDS-PAGE

fluo-rography (23); fractionswere pooled andused asthe IL-1 monocyte

standard.

hrIL-1. The IL-IcDNAexpressionvectorwasconstructedby isolating

the 1112bp NcoI-XmnIfragment(bp295-1407)from the IL-1 cDNA plasmidpcD12-1-8 (19) andinserting it intoanE.coliexpression plas-mid. The hrIL-l is 223 amino acids long and contains 199 of the 269 amino acid longIL- 1 precursor.Theaminoterminal 24 amino acidsare

contributed by theexpression vector. ThehrIL- I included 46 amino acidsof theIL-Iprecursorpeptide, whichare presentbefore the alanine

atposition 117. This alanine istheN-terminusoftheprocessedIL-I

found in thesupernatesof stimulated human bloodmonocytes(24).The predictedmolecularweight of the hrIL-l is 24.5.hrIL-l-producing bac-teriaweregrown at37°C, lysed, and thehrIL-1wasextracted from the insolublecellular fraction with 8M urea. hrIL- I waspurified by sequential ion-exchange and either gel-filtration chromatography orhigh

perfor-manceliquid chromatography. The hrIL-l was storedat -70°C ina

0.15Mphosphate-bufferedsaline,pH 6.8. Theidentityof thepurified

hrIL- 1 wasconfirmed by amino acid composition and thesequenceof theaminoterminalheptapeptide.Protein concentrationwasdetermined by the Bradford methodusingabovineserumalbumin (BSA) standard. PuritywasassessedbySDS-PAGE with Coomassie Blueorsilverstaining

(Fig. 1). The hrIL-l is estimatedtobe >99%pure.Theendotoxin

con-centrationof thehrIL- I was -20pg/jigofIL- I protein.

Insubsequentexperiments forpyrogenassays inrabbits,41amino acids ofthe precursorpeptidewere removedbyenzymatic digestion,

which yielded hrIL-l withanN-terminusat the alanine position 112

and amolecularweightof 17,500. This form of IL-I is similartothe

maturepeptidepurified frommonocyteswith theN-terminusatalanine position 117 (24). This recombinant form ofIL- Ifrom amino acid

112-269washomogeneousonSDS-PAGEandcontained 60pgof endotoxin/

mgof IL-l protein.

LAF assay. For thecostimulatorassay,theclonedmurine helperT

cellline,D.10.4G.1 wasused(25). Previous studies revealed that thisT

cell line is sensitivetohumanmonocyte IL- 1 preparations(26). Dilutions of IL-1 weremade in RPMIcontaining5% heat-inactivated fetal calf

serum(FCS)and25

Mg/ml

ofpolymyxinB. Fortheassay,100,1of each dilutionwasadded toflat-bottom microtiter wells intriplicate. TheT

cellswereused 10-14dafter the last feeding, washed and suspended in

5% FCS/RPMIat 2 x IO'cells/ml. All wellsreceived 100 Ml ofthe T cellsuspension containing 1

Mg

PHA/ml. Theproliferativeresponsesof these cellsweremeasuredby uptake of tritiatedthymidine after72h.

HumanTcell responses. Humanperipheralblood T cells werepurified bya combination ofadherence toplasticandnylon wool. Tcells (5

X 106/ml)wereincubated for3 d inflat-bottomwells in the presenceof

variousconcentrations ofhrIL-l and/orPHA(0.5,g/ml). During the last 12h, cellswerepulsedwith[3H]thymidineand then harvested.For IL-2production, cells were incubated in round-bottom culture tubes for 3din 10 mMhydroxyureaand the presence or absence of hrIL-I and

0.5 ,g/mlPHA. Thesupernatantswere_dialyzed andassayedforIL-2

t I s .

I

ai

&w

J;=l

Figure1. SDS-PAGE ofhrIL-I used in these studies. The gel

wasloaded with the crudeurea

extract from the recombinant

IL-I strain(lanesI and4). Pu-rifiedhrIL-l (50 Mg) is shown inlane3.Lane2 contains

1

2

3

4

marker proteins.

usinganassay based uponincorporation of[3H]thymidinein IL-2-de-pendent PHA-activated cultured human Tlymphocytes (27). Control cultureswerestimulated with known concentrations of recombinant IL-2(Cetus Corp.,Berkeley, CA).2-d supernatantswerealsoassayedfor IL-2usinganIL-2-dependentcytotoxicTlymphocyteline(CTLL).All errorbars shown infiguresrepresentmeancpm±SD.

Pyrogenassays.Amodification of themousepyrogen assay(28)was

used.C3H/HeJmice(JacksonLaboratories,BarHarbor, ME)were im-plantedsubcutaneously withradiotransmittingthermosensors (Minim-itter, Sun River, OR). Each thermosensorwascalibratedpriorto

im-plantation. Temperatureswere recorded every 10 minusing adigital

analogue offrequencyintervalsreceived fromanAMradio(29).Mice wereplaced ina34°C infant isolette for 1.5 h and baseline temperature

wascalculatedduring60min beforeintraperitoneal injection.Rabbits (New Zealand-derived albinofemales) weighing2.5kgwere used for pyrogentestingaspreviouslydescribed(23). Temperatureswererecorded every minuteusingaKayeInstrumentsDigistripIIrecorder(Bedford,

MA). The data represent the first dayresponses of trained rabbitsto

hrIL- 1.Threerabbitsgiven 1Mg/kgwereusedinthegroup that received 10Mg/kg.Rabbitsweregiven dailyinjectionsofhrIL- Iin ordertoinduce pyrogenic tolerance; the day 2-5 responses of these rabbitswere notused

tocalculatethemeanfevers.

Dermalfibroblast PGE2 production.Newborn human foreskin fibro-blast cultureswereseededonto2-ml flat-bottom wells(Costar,Cambridge,

MA) and growntoconfluence inRPMIcontaining5%FCSaspreviously

described(30).24 hbeforetheaddition ofIL-1,cellswerewashed and incubatedovernightinRPMIwithoutserum. IL- Idilutionsweremade inserum-free RPMI andwereusedtoreplacethe overnightmedium. After24h, fibroblastsupernateswere removed andfrozenat -70°C.

Prostaglandin E2 (PGE2) concentration was determined by radioim-munoassay(Seragen Inc., Boston, MA).

Results

Activity ofhrIL-J inthemurine LAF assay. The LAFactivity oftheurea extractsfromthe recombinantE.colicontainingthe IL- I insertwascomparedto urea extractsfrom recombinantE.

colicontainingtheexpression plasmidbutwithout theIL- I

se-quence.Theproteincontentof thetwoextracts wasadjustedto

the same concentration. When diluted, the extracts from the recombinantIL-I-containing strain had considerably more

1:100,000 vs. 8.3, 1.7, and 1.2, respectively). During purification, LAF activity corresponded to the intensity of the 24,000-mol wt protein staining in SDS-PAGE. The homogeneous hrIL-l was tested on the murine T cell line,DlO.G4. 1, but there was noproliferative responses to hrIL-1 alone(100 ng/ml); however, in the presence of a suboptimal concentration of PHA, marked increases in thymidine uptake occurred (Fig. 2). At a concen-tration of 0.05 ng/ml, hrIL-1 doubled the PHA response (P <0.05, Student's paired ttest). Using half-maximal responses

asthe definition of an IL-1unit, the hrIL-1has aspecific activity of 5 X 106 U/mg protein and is comparable to that reported for recombinant murine IL-1 (6 X 106 U/mg) (21). Fig. 2 A illustrates

1of more than 20 separate assays using these T cells as indicators of hrIL-1 activity and in eachexperiment similar specific activ-ities were observed. As shown in Fig. 2 B, antibody raised to human monocyte IL-1 (23) neutralized the recombinant IL-1.

Inaddition, the recombinant IL-1 molecule, like the monocyte-derived IL-1, is heat labile; heatinghrIL- 1 for 30 min at 70°C reducedactivity by 75% (data not shown).

HumanTcell responses.Two different assays were employed usinghuman Tcells stimulated in the presence of hrIL-1 and mitogen. A human T cell blastogenesis experiment is shown in Fig. 3 A that illustrates that hrIL-1 in the absence of PHA has noeffect on T cell proliferation. In the presence of PHA, however, hrIL-1 atconcentrations of 100 and 10 ng/ml induced human

Tcellproliferative responses. In another assay, human peripheral bloodTcell IL-2production was measured during a 3-d incu-bation. During the incubation period, hydroxyurea was added

toreduce IL-2binding to theTcells. Supernatant and control mediaweredialyzed and assayed on IL-2-sensitive T cells. As

A

15-I 10-x

E

,5

PHA 0.05 0.5 5 50 500

hrIL-1 (ng/ml)+PHA

B

I*

° 20-x

E

C,

-

10-C.)

0

jl

NRS 0.1% NRS0.01% Anti-L-1 Anti-L-1

Figure 2. (A) Proliferative responsesofDlO.G4. 1 cells tohrIL-l in the presenceofPHA.Concentration ofhrIL-1 isindicatedunder the

hori-zontalaxis. (B) Effect ofantimonocyte IL-l onhrIL-l.hrIL-l (5 ng/

ml)wasincubated witheither normalrabbit serum (NRS) oranti-IL-l

atthe concentrations (vol/vol)indicated for2 h at37°Cbeforebeing

addedtotheDIO.G4.1 cells. PHA=2,533±322cpm; the responseto

5 ng/ml of hrIL-l withorwithoutthe NRSwaswithin10%.

shown inFig. 3 B, hrIL-1 didnotinduceIL-2 unless PHAwas

present;significant IL-2 production occurredatIL-I

concentra-tions of 5ng/ml.

Additional experimentswerecarriedoutin which human T cellsdepleted ofmonocytesby adherence followed bytwo pas-sages over nylon wool. These nonadherent cells (T cells)were

then incubated with PHA in thepresenceof various

concentra-tions of hrIL-1 (as above) for 2 d and the supernatant medium dialyzed beforeassay.These supernatantswereassayedonCTLL cells for thepresenceof IL-2. In Fig. 4,weshow that under these conditions, hrIL-1 in thepresenceof PHA induced IL-2 from nonadherent(NA) cellsatlevels comparabletothosefrom

un-fractionated mononuclear cells incubated with the mitogen. NA

cells stimulated with mitogenin the absence of IL- 1 produced significantly less IL-2 but this levelwasseveralfoldgreaterthan CTLLresponses to unconditionedmedium (cpm = 346±52).

FibroblastPGE2production. hrIL-1 isa potent inducerof PGE2 synthesis in human foreskin fibroblasts.AsshowninFig. 5,hrIL- I stimulated dermal fibroblast PGE2 productionat

con-centrations of 5-500 ng/ml. Concon-centrationsof0.5ng/ml induced small butsignificant levels (P< 0.05). The fibroblasts donot

produce PGE2under the same conditions when incubated24h

inthe presence of as much as 1

_Ag/ml

of endotoxin (30). A

x

E

Q

0

cr

-In-Medium PHA 100ng 10 ng 100 ng 10 ng

PHA PHA

Conc. hrlL-1 (perml)

B

0 x

E

c.) 0

._

C.)

N -J

Q hrIL- 1 (3d)

*hr IL-1+PHA (3d)

Ehr IL- 1+PHA(Od) *p<0.01

Medium 500ng/ml 50ng/mI 5ng/mI

Figure 3. (A) Human T cellblastogenicresponsetohrIL-land PHA. Responsesof unfractionatedcellsto0.5gg/mlPHAwere9,533±76.

ConcentrationofhrIL-1 is indicatedunder thehorizontal axis.(B)

ProductionofIL-2fromhuman T-cells.SupernateswereassayedonT

lymphoblasts (27)inquadruplicate.The mixtureofPHAplushrIL- I wasdirectlyassayedwithoutprior incubation withTcells(indicated

as0d)andtheamountofinsituIL-2generationis reflectedbythe difference in 3 d and 0 d. The asterisks indicate *, level ofsignificance

between the valuesat3 dand0 dusingapairedttest.

A

06

MNC+PHA NA NA+PHA 100 10 1

NA+PHA+hrIL-1 (ng/mI)

Figure 4. Generation of IL-2 fromNAcells stimulated withhrIL-l

and PHA. NA wereprepared byadherenceandtwopassagesover ny-lon wool and incubated with or withoutPHA(see Methods). After 48 hthesupernatantswereassayed for IL-2usingCTLL cells in

quadru-plicate. Thedata show a dilution ofsupernatantsof 1: 16.

Fever in mice. Endotoxin-resistant mice (C3H/HeJ) were

used in thesestudies. To ruleoutendotoxincontamination,the

concentrationin the purified hrIL-Iwasdeterminedby alimulus amebocyte lysateassaythat hadasensitivity of 20 pg/ml.

End-point determinations using the clot-forming method revealed

that hrIL-1 contained 20 pg/ag of IL-I protein. C3H/HeJ mice used undertheconditions of thisstudy developfever (maximum

temperature,Tmax =0.4°C)after the injection of500 ng per 25-g mouseofE.coliendotoxin. As shownin Fig.6A, the hrIL- 1

inducesabrisk, monophasic fever inprewarmed mice. Heating

thesample resulted in loss of pyrogenicity. Thedecreaseinbody temperaturethat occurredaftertheinjection of heat-inactivated

hrIL- 1 is frequently observed in studiesusing prewarmed mice orratsfollowingtheinjectionof nonpyrogenicagents.Thismay

be, inpart, dueto removalof the animals from their

thermo-neutralenvironment during the injectionprocedure. The dose response for the pyrogenic response is illustrated inFig. 6 B.

Similar to previous studies on human monocyte IL- 1 in the

murinefevermodel, increasing the dose ofpyrogen often

pro-duceslessfever(28).

CO

cu

1-20 U.

o0.5

x

0)

0.

Medium Monocyte 500 50 5 0.5 IL-1

hrlL-1 (ng/ml)

Figure 5. Human dermal fibroblast PGE2 production.hrIL-l was

in-cubatedfor24h induplicatewellsandPGE2levelsdeterminedas

de-scribed(30).

0

'1

0.5 0.4 0.3 0.2

0.1

.

.

I

IX

10 20 30 40 50 60 70 85 100

\ ^he~~~~~~atedtr>7)

\X < ~~~70'Cx20min

B 0.5

e30.4 -i6

., 0.3

-

0.2-0.1 2A,ug/kg

(70oCx2o min)

15 2 4

1

.5

fig/kg f

Figure 6. hrIL-l fever in C3H/HeJ mice. (A)Meanfevercurvein miceinjected with hrIL- 1. (B) Dose-response of fever in micegiven

hrIL-l.

Fever inrabbits. Rabbits injected intravenously with

hrIL-1 (amino acids 71-269) developed typical monophasic fevers thatreachedpeaktemperatureelevations -54 min after

intra-venousinjection (Fig.7).Preincubation withpolymyxinB(250

jig/ml)

inhibits the pyrogenic property ofendotoxins but did

notreduce the fever produced byhrIL- 1.Heating for30 minat

70°C,aprocedure that doesnotaffect thebiological properties of bacterial endotoxins, results in loss ofpyrogenicity ofIL-1 derived frommammaliancells (5,9, 12)and abolished the

pyr-ogenicityof hrIL-1. Inotherexperiments, 5 dailyinjectionsof

hrIL-1 inrabbitsdidnotinducepyrogenictolerance (datanot

shown).

0.8-

0.6-o0.4

lOuglkg,PmxB

0.2-

-0

(n.6)M70C,30min)

0 30 60 90 120 190 180 Minutes

Figure 7. hrIL- 1 feverinrabbits.Meanfeversfollowingintravenous injections of hrIL-l (24,500molwt) intorabbits. The numbersof

rab-bits ineach groupareindicatedinparentheses.Heated hrIL-1 wasnot

Inadditional fever experiments, the hrIL- 1injectedwasthe shorter form (amino acids 112-269, 17,500 mol wt), which con-tained -60 pg of endotoxin/mg of IL-I protein. This hrIL-l was50-fold more pyrogenic in rabbits than the larger hrIL- 1 (molwt 24,500) containing the 46 precursor amino acids (com-pare Fig. 7 with Fig. 8). The murineTcell responses were sim-ilarlyincreased 50-fold.

Discussion

Theobject of these studies was to establish that homogeneous preparationsof human IL- 1 purified from a recombinant strain of E.coli are biologically active in assays of both immunological function andinflammatory responses. Although thehrIL-1 used inthese studies will need to be further evaluated for its ability to induce physiologic and metabolic changes associated with acute-phase responses, the data presented in this report support theconcept that recombinant human IL- I functions as afactor thatpromotesmitogen-driven T-cell blastogenesis and IL-2 pro-duction as well as induces fever and acts as a mediator of in-flammation. Furthermore, these properties are destroyed by mild heating, which is characteristic of monocyte-derived IL-1.

The human IL- I encoded by the precursor cDNA is unique in that it lacks a clear signal peptide that would indicate a likely processing site for the N-terminus. A murine IL-IcDNA is sim-ilar inthat it encodes for a similar size precursor and that it also lacks a signal peptide. Recent studies of the N-terminus amino acid sequence of human monocyte-derived IL-I witha plof7

haveindicated that the cleavage of the precursor at the alanine (amino acid 117) is the likely N-terminus of the IL-I found in supernatesofstimulated monocytes (24). The IL-I expressed in therecombinant E. coli and tested in the present study contains 46of the 1 16 amino acids that are part of the precursor peptide. Several investigators have isolated peptides with molecular weights between 38,000 and 30,000 employing gel-filtration chromatography, and which exhibit many of the same IL- I

bio-1.0

0.8

0.6-

0.4-0

0.2

-0 30 60 90 120 150

TIME(MINUTES)

Figure8. hrIL- I fever in rabbits. Individual rabbit responses to hrIL- I (mol wt 17,500)followinganintravenous injection of 500ng/kg are

shown.PolymyxinB orproteinstabilizerswere notincluded.

logical

activities reported for the 15,000-17,000-molwtsize (12-14, 31, 32). Although some reports indicate that the human

IL-I pl7form without cleavageofthe precursor peptide haslittle biological activity (20), the present study establishes that the proximal 46amino acids of the precursor peptide do not prevent biologicalactivity in the assays used in these studies. This ob-servation supports those of other investigators who have reported multiple biological activities of large molecular weight IL-Is. Since hrIL- I functions indistinguishably from that of monocyte IL-1, we conclude that the extra amino acids derived from the precursor sequence do not prevent biological activity.

Theseadditionalamino acids may affect biological potency ofthe material, however. Nevertheless, the specific activity of the hrIL-l fromE.coli (5X 106half-maximal units/mg) is nearly the same as thatfor recombinant murine IL- 1 (21). The murine recombinant IL-I expressed in that study did not contain any precursor sequences. A recent report on the specific activity of purified human monocyte IL-I with a pl of 7 indicates a specific activity 4.3 X I07 U/mg protein (8). By removing all but five of the precursoramino acids (N-terminus at alanine position 1 12 comparedwith the natural N-terminus at alanine position 1 17), we observed a 50-fold increase in the specific activity of the hrIL-I for both murineTcell stimulation as well as in pyrogen-icity in rabbits. Cleaving the precursor peptides and producing hrIL-1 with a N-terminus closer to that of the naturalIL-I clearly increased the activities in these models but the shorter hrIL- 1 did not have a significantly increased specific activityondermal fibroblasts (data not shown). This may be due tothe fact that some cell cultureconditionsprovide sufficient proteasestocleave theprecursoraminoacids;alternatively,the precursor polypep-tides may affect the potency of IL- 1 for certain cells.

The specific activity of hrIL- I may be affected by several conditions including the denaturization during extraction and purification, the additional amino terminal amino acids, and the lack ofglycosylation. To isolate the IL-1 from the recom-binant E. coli, denaturing conditions were employed. Refolding of the molecule may not be complete or isolation procedures may partially inactivate sites critical for biological activity. The lack of glycosylation by E. coli may influence the specific activity orsolubility of hrIL- 1. The peptide sequence of theIL- I indicates a likelyglycosylation site but the nonglycosylated form used in thesestudies establishes that the immunologic andinflammatory properties of IL- 1 are maintained. Murphy was unable to show that the charge heterogeneity of rabbit IL- Is was due to glyco-sylated variants (33). Studies presently underway using

mam-malian cell-derived recombinant IL- I will be able toclarify this issue further.

Manybiological properties ofIL-1 are sharedwith bacterial endotoxins. It is thus important to exclude endotoxin from any IL- 1 preparation, particularly recombinant IL-I produced inE.

coli. Although the preparations used in the present study had lowendotoxin content (20pg/,ug)asdetermined by the Limulus amebocyte lysate assay, we used several methodstodemonstrate that the biological effectsofhrIL-1 were not duetoany unde-tected endotoxin. The fibroblast PGE2 stimulation assay (30) andC3H/HeJ micearehighly insensitivetoendotoxins.Heating hrIL- 1 to 70°C destroyed itspyrogenicity inmice and rabbits and polymyxin Bhad noeffect on hrIL- 1. Nevertheless, endo-toxins are potent pyrogens and produce small fevers

(minimum

pyrogen dose)inrabbits at 3ng/kg.TheshorterhrIL-1 (molwt

thatwould beequivalentto30fg endotoxin/kg. Thus,therecan

nolonger be any question thatIL- 1 is anendogenouspyrogen

andthatitspyrogenic propertiesare not due to contaminating endotoxins. These studiesclearlydemonstrate that hrIL- 1 pos-sessesmultiplebiologicactivities attributed topurified material obtained from blood monocytes and that the properties ofT cell activation and induction ofinflammatory responsesreside

inasingle polypeptide.

Noteaddedinproof INadditional feverexperiments,we injectedthe 17,500-mol wthrIL-I into 6rabbits and observed a meanfeverpeak

of 0.68±0.07°Cfollowingadose of 100 ng/kg.

Acknowledgments

The authors thank Walter0. Fredericks of Cistron Technology,Inc.and

Thomas Fraserof Repligen Corp.for theirencouragementof these studies. Specialthanks to Dr.TakashiIkejima,ScottOroncle, Redentor Maxwell,

andLorraine Barrie for assistance.

Supported byNational Institutes of Healthgrants Al 15614 and CA 39489 fromthe U.S. Public Health Service andby CistronTechnology, Inc., PineBrook, NJ.

References

1.Nathan, S. B., H. W. Murray, and Z. A. Cohn. 1980. The

mac-rophageas aneffectorcell. N. Engl.J.Med. 303:622-626.

2.Dinarello,C. A.1984. Interleukin-1.Rev.Infect.Dis. 6:51-95.

3.Dinarello,C. A. 1984.Interleukin-I andthepathogenesis ofthe acutephaseresponse. N.Engl.J.Med. 311:1413-1418.

4.Krane, S. M.,S.R.Goldring,and J.-M. Dayer. 1982. Interactions

amonglymphocytes, monocytes, and othersynovialcellsinthe

rheu-matoidsynovium.InLymphokines.E.Pick andM.Lany,editors.

Ac-ademicPress, New York. 75-136.

5.Rosenwasser,L. J., C. A.Dinarello,and A.S. Rosenthal. 1979.

Adherent cellfunction in murine T-lymphocyte antigen recognition.IV.

Enhancement ofmurineT-cell antigenrecognitionby humanleukocytic

pyrogen. J. Exp.Med. 150:709-714.

6. Murphy, P. A., P. L.Simon, and W. F.Willoughby. 1980.

En-dogenouspyrogensmadeby rabbitperitonealexudatecellsareidentical withlymphocyteactivating factorsmade byrabbitalveolarmacrophages. J.Immunol. 124:2498-2501.

7.Kampschmidt,R. F. 1984.Infection, inflammation and interleu-kin-1. LymphokineRes. 2:97-110.

8.Matsushima, K., S.K.Durum, E.S.Kimball,and J. J.Oppenheim. 1985.Purification ofhumaninterleukin-1 andidentity ofthymocyte

co-mitogenic factor, fibroblast proliferation factor,acute-phase protein-in-ducing factorandendogenouspyrogen.Cell.Immunol. 92:290-302.

9. Hanson,D. F.,andP. A.Murphy. 1984.Demonstration of inter-leukin-I activity in apparently homogeneous specimens ofthe pI 54

form of rabbit endogenouspyrogen.Infect.Immun. 45:483-490.

10.Dinarello, C.A., H. A.Bernheim,J. G. Cannon,G. LoPreste,

S.J.C.Warner,A.C. Webb, andP. E. Auron.1985.Purified,"S-met, 3H-leu-labeledhumanmonocyteinterleukin- 1 with endogenouspyrogen

activity.Br. J.Rheumatol. 24:59-64.

11. Lachman, L. B. 1983. Human interleukin-1: purification and

properties. Fed.Proc.42:2639-2644.

12.Dinarello, C. A., N. P. Goldin, and S. M. Wolff. 1974.

Dem-onstrationandcharacterization oftwodistincthumanleukocytic pyr-ogens.J. Exp.Med. 139:1369-1381.

13. Wood, D. D., E. K. Bayne, M. B. Goldring, M. Gowen, D. Ham-erman, J. L.Humes, E. J.Ihrie,P. E.Lipsky,and M.-J. Staruch. 1985. Thefour biochemically distinctspecies of humaninterleukin-l allexhibit

similar biologic activities.J.Immunol. 134:895-902.

14. Sahasrabuddhe, C. G., C. A. Dinarello, B. Martin,and A. L.

Maizel. 1985.Intracellular humaninterleukin-1:aprecursor for the se-creted monokine. Lymphokine Res. 4:205-213.

15. Saklatvala, J., J. Pilsworth, S. J. Sarsfield, J. Gavrilovic, and

J. K.Heath. 1984.Pig catabolin isaform of interleukin-l.Biochem.J. 224:461-466.

16.Kimball,E.S., S. F.Pikeral,J.J.Oppenheim, andJ. L.Rossio. 1984.Interleukin-Iactivityinnormalurine. J.Immunol. 133:256-261.

17. Dinarello, C. A.,G. H. A. Clowes, Jr., A. H. Gordon, C. A.

Saravis,Sand S. M. Wolff. 1984.Cleavageofhumaninterleukin-l:

iso-lation of a peptide fragment from plasma of febrile humansand activated

monocytes. J.Immunol. 133:1332-1338.

18. Cannon,J.G., and C.A.Dinarello. 1985. Increasedplasma

in-terleukin- 1 activityin women after ovulation. Science(Wash.DC).227: 1247-1249.

19. Auron, P. E., A. C. Webb, L.J. Rosenwasser, S. F. Mucci,A.

Rich, S. M.Wolff, and C. A. Dinarello. 1984. Nucleotide sequenceof humanmonocyteinterleukin-1 precursor cDNA. Proc.Natl. Acad.Sci. USA. 81:7907-791 1.

20. March, C. J., B. Mosley, A. Larsen, D. P.Cerretti, G.Braedt, V.

Price, S. Gillis, C. S. Henney, S. R. Kronheim,K.Grabstein,P. J.Conlon,

T. P. Hopp, and D.Cosman. 1985.Cloning,sequenceandexpression

of two distinct human interleukin-1 complementary DNAs. Nature

(Lond.). 315:641-646.

21. Lomedico, P. T., U. Gubler, C. P. Hellmann, M. Dukovich,

J.G. Giri, Y.-C. E. Pan, K. Collier, R. Semionow, A. 0.Chua, and

S.B.Mizel. 1984. Cloning and expression of murine interleukin-I cDNA

in Escherichia coli. Nature(Lond.). 312:458-462.

22. Webb, A. C., P. E. Auron, A. Rich, L. J.Rosenwasser, S. M. Wolff, andC. A. Dinarello. 1985. Molecular cloning of human

interleu-kin-1 precursor cDNA anditsexpression in monkey cells.In:Molecular

andCellular BiologyofLymphokines. IV. International Lymphokine Conference.C.Sorgand A.Schimpl,editors. AcademicPress, NewYork. 685-695.

23.Dinarello, C.A., L.Renfer,andS.M.Wolff. 1977. Human leu-kocytic pyrogen: purification and developmentofaradioimmunoassay.

Proc. Natl.Acad. Sci. USA. 74:4624-4627.

24. Van Damme, J., M. De Ley,G.Opendenakker,A.Billiau,and P.DeSomer. 1985.Homogeneousinterferon-inducing 22K factor is re-lated to endogenous pyrogen and interleukin-1. Nature(Lond.). 314:

266-268.

25. Kaye, J., S. Porcelli,J.Fite,B. Jones, and C.A.Janeway, Jr.

1983. Both a monoclonalantibodyanantisera specific fordeterminants

uniquetoindividual cloned helper T-cell linescansubstitute for antigen

andantigen-presenting cells intheactivation of T-cells.J. Exp. Med.

158:836-843.

26. Kaye, J., S. Gillis, S. B. Mizel, E. M. Shevach, T. R. Malek, C.A.Dinarello,L.B.Lachman,andC.A.Janeway,Jr. 1984. Growth ofaclonedhelperT-cellline induced byamonoclonalantibody specific for the antigenreceptor:interleukin-I isrequired fortheexpression of

receptorsfor interleukin-2. J. Immunol. 133:1339-1344.

27.Gillis,S., M. Ferm, W. Ou, and K. Simth. 1978.Tcellgrowth factor:parametersofproduction andaquantitativemicroassay for

ac-tivity. J. Immunol. 120:2027-2031.

28.Bodel, P., and H.Miller. 1978. A newsensitivemethodfor

de-tecting human endogenouspyrogen.Inflammation. 3:103-110. 29.Cannon,J.G., and C. A.Dinarello. 1985.Multiple

interleukin-1 activities in lutealphase humanplasma. Br. J.Rheumatol. 24:226-229.

30.Bernheim, H. A., andC.A.Dinarello. 1985.Effects of purified

humaninterleukin- 1 on the release of prostaglandin E2 from fibroblasts. Br.J.Rheumatol. 24:122-127.

31. Dinarello, C. A., K. Bendtzen, and S. M. Wolff. 1982. Studies

ontheactive site ofhumanleukocyticpyrogen.Inflammation.6:63-78. 32.Kampschmidt,R. F.,and L. A.Pulliam. 1978.Effect ofhuman monocytepyrogen on plasma iron, plasmazinc,and bloodneutrophils.

Proc.Soc. Exp. Biol.Med. 158:32-35.

33.Murphy, P. A., T. A. Cebula, and B. E. Windle. 1981. Hetero-geneity of rabbitendogenous pyrogen is not attributable toglycoslylated