IL-1Ra Danielle Burger and Jean-Michel Dayer * Division of Immunology and Allergy, Department of Internal Medicine, University Hospital, CH-1211, Geneva 14, Switzerland *corresponding author tel: 41-22-372-9376, fax: 41-22-372-9369, e-mail: [email protected] DOI: 10.1006/rwcy.2000.04002. SUMMARY in numerous experimental models of diseases in animals. Because of its beneficial effects in many animaldiseasemodels,IL-1Rahasbeenusedasthera- Inflammation is an important homeostatic mecha- peutic agent in human patients. IL-1Ra has failed to nism that limits the effects of infectious agents. show beneficial effects in septic shock but seems to However, inflammation might be self-damaging and therefore has to be tightly controlled or even abol- improvetheconditionofrheumatoidarthritispatients. ishedbytheorganism.Interleukin1(IL-1)isacritical mediator of the inflammatory response, playing an BACKGROUND important part in the development of pathologic conditionsleadingtochronicinflammation.Although IL-1 production may be downmodulated or its effect Interleukin 1 receptor antagonist (IL-1Ra) is a limited by so-called anti-inflammatory cytokines, member of the IL-1 family. Three forms of IL-1Ra IL-1 inflammatory effects are inhibited and can be have been described, two of them being expressed as abolished(invitro)byoneparticularlypowerfulinhib- intracellular proteins (icIL-1RaI and icIL-1RaII) and itor, IL-1 receptor antagonist (IL-1Ra). one being secreted (sIL-1Ra). The three forms of IL- To date, three forms of IL-1Ra have been iden- 1Raresultfromthesamegene.Thefunction(s)ofthe tified, one soluble form and two intracellular forms, intracellular forms of IL-1Ra are still elusive whereas referred to as sIL-1Ra, icIL-1RaI, and icIL-1RaII, sIL-1Ra binds competitively to IL-1 receptor I respectively. Although sIL-1Ra functions obviously without inducing signal transduction and thus inhib- as an IL-1 inhibitor by competitively binding to IL-1 its IL-1(cid:11) and IL-1(cid:12) actions. receptor without inducing signal transduction, the functions of icIL-1RaI and icIL-1RaII remain to be Discovery determined. The three forms of IL-1Ra are tran- scribed from the same gene, two different promoters regulating the transcription of sIL-1Ra and icIL- An IL-1 inhibitory activity was first observed in the 1RaI. In mice, sIL-1Ra is predominantly found in urine of febrile patients (Balavoine et al., 1986). peripheral blood cells, the lung, spleen, and liver, Subsequently this activity was further attributed to while icIL-1RaI is mainly found in the skin. IL-1Ra sIL-1Ra, which blocks IL-1 binding to its receptor knockout mice display growth retardation after (Seckinger et al., 1987). sIL-1Ra cDNA was cloned weaningandanincreasedsusceptibilitytoendotoxin- from a human monocyte library (Eisenberg et al., induced injury and collagen-induced arthritis, sug- 1990).Thereafter,sIL-1Rawasfoundtobeproduced gesting that IL-1Ra plays an important part in both bymanyothercellsincludingneutrophils,fibroblasts, healthandpathologicconditions.IL-1Raisproduced corneal epithelial cells, and microglial cells. Indeed, by hepatic cells as an acute phase protein. The sIL-1Ra production is inducible in most cell types. circulating blood levels of sIL-1Ra are low in normal Natural sIL-1Ra is a 22–26kDa glycoprotein conditionandelevatedinindividualswithallele2poly- (Seckinger et al., 1987; Hannum et al., 1990; Mazzei morphism and in several human infectious, auto- et al., 1990), whose recombinant 17kDa form retains immune,andchronicinflammatorydiseases,aswellas full IL-1-inhibitory capacity (Carter et al., 1990; 320 Danielle Burger and Jean-Michel Dayer Eisenberg et al., 1990). Since it potently inhibits the sIL-1Ra sequence was conserved during evolution, various effects of IL-1, sIL-1Ra is considered an displaying >75% homology in human, rat, rabbit, important regulator of the inflammatory and overall mouse, bovine, and horse sIL-1Ra. sIL-1Ra interacts immune response mediated by IL-1. sIL-1Ra was with the same receptors as IL-1(cid:11) and IL-1(cid:12) soon regarded as a potential marker of disease suggestingstructuralratherthansequencesimilarities severity (Prieur et al., 1987) and a therapeutic tool between these cytokines. It has been proposed that (Seckinger et al., 1990). icIL-1RaI was also cloned sIL-1Ra binds to the IL-1 receptor I through one fromahumanmonocytelibrary(Haskilletal.,1991). interacting domain as compared with two such NaturalandrecombinanticIL-1RaIbindstotheIL-1 domains on IL-1(cid:11) and IL-1(cid:12). The lack of a receptor I with a similar affinity to sIL-1Ra in vitro receptor-interacting domain might be the cause of (Gruaz-Chatellard et al., 1991; Haskill et al., 1991). the lack of signal transduction upon binding of sIL- icIL-1RaI is constitutively expressed in keratinocytes 1RatoIL-1receptorI(forreviewseeLennard,1995). and epithelial cells (Arend, 1993; Arend et al., 1998). Another putative intracellular form of IL-1Ra was Main activities and identified by polymerase chain reaction on mRNA pathophysiological roles from human polymorphonuclear cells (Muzio et al., 1995). However, the protein product of this high sIL-1Ra competitively binds the IL-1 receptor I molecular weight icIL-1Ra mRNA form has never without inducing signal transduction. It is therefore been detected. More recently, a 16kDa icIL-1Ra was likely that sIL-1Ra regulates (inhibits) cellular detected by western blot analysis of LPS-stimulated functions affected by IL-1(cid:11) and/or IL-1(cid:12). The humanneutrophils,monocytes,andthehumanhepa- function(s) of intracellular forms of IL-1Ra are less toma cell line HepG2 (Gabay et al., 1997b; Malyak obvious. Some hypotheses were recently reviewed et al., 1998). This low molecular weight form of icIL- (Arend et al., 1998) claiming that icIL-1Ra might 1Ra is now referred to as icIL-1RaII. Furthermore, counteract intracellular IL-1(cid:11) activities and destab- a variant cDNA species containing an additional 171 ilize and/or degrade mRNAs induced by IL-1(cid:12). nucleotides within the icIL-1RaI cDNA, that inter- However the latter functions remain to be demon- rupts the coding region, was recently described strated. More recently, it has been shown that resting (Weissbach et al., 1998). This mRNA variant was and cytokine-stimulated human pulmonary epithelial found to be expressed in keratinocytes and human cells release icIL-1RaI in the extracellular space, articularcartilage. Translation is likely tobe initiated where it can antagonize cell surface IL-1R (Levine atanalternateMetcodonotherthanthatutilizedfor etal.,1997).Therefore,icIL-1Ramightbeanepithelial icIL-1RaI,suggestingthatthismRNAvariantencodes store of IL-1Ra liable to immediate release upon a novel polypeptide. The function(s) of icIL-1Ra iso- stimulation. Furthermore, the expression of icIL- forms remain(s) unclear. 1RaI in Caco-2 cells decreased IL-1(cid:12)-induced IL-8 secretion (Bocker et al., 1998). Whether icIL-1RaI is Alternative names released by the latter cells remains to be determined. IL-1Ra was also known as IL-1 receptor antagonist GENE AND GENE REGULATION protein (IRAP), but it is now currently referred to as IL-1Ra or sIL-1Ra (secreted IL-1Ra) and its intra- Accession numbers cellular forms as icIL-1RaI and icIL-1RaII. The IL-1Ra gene is referred to as IL-1RN. The Structure sequence of human IL-1RN, bases 1–33,414, is accessible in GenBank at number U65590. cDNA sequences are accessible at number M63099 (human Mature human sIL-1Ra is a 152 amino acid residue sIL-1Ra) and X84348 (human icIL-1RaI) and protein which is generated by the cleavage of a 25 X52015, X64532, X53296, and M55646. Mouse IL- amino acid hydrophobic signal sequence from a 1RN is accessible at number MGI:92197. cytoplasmic precursor of 177 amino acid residues. Although IL-1Ra, IL-1(cid:11), and IL-1(cid:12) genes seem to Chromosome location have arisen through duplication and divergence of a common ancestral gene, sIL-1Ra displays no more than 19% and 26% amino acid sequence homology Human IL-1RN has been mapped in the long arm of with IL-1(cid:11) and IL-1(cid:12), respectively. However, chromosome 2 at bands q14-21 (Steinkasserer et al., IL-1Ra 321 1993). Mouse IL-1RN has also been mapped at one minor and two major sites, 14, 15, and 16 chromosome 2 (10.00 cM) (Zahedi et al., 1991). nucleotides, respectively, upstream to the cDNA sequence. icIL-1RaII is derived from both icIL-1RaI and sIL-1Ra mRNA by alternative translation Relevant linkages initiation from the second 50 ATG (Malyak et al., 1998). HumanIL-1(cid:11)andIL-1(cid:12)genes(IL-1AandIL-1B)are A variable number of tandem repeats polymorph- also located in the long arm of chromosome 2. There ismhasbeendescribedinintron2oftheIL-1Ragene are only 75kb between IL-1A and IL-1B and 200kb (Tarlowetal.,1993).Allele2ofthispolymorphismis between IL-1B and IL-1RN. The genes of the associated with the severity of many chronic inflam- mouse IL-1 family are mapped in chromosome 2 matorydiseases(Table1),althoughthemechanismof but IL-1RN is far from IL-1A and IL-1B which are this association remains elusive. Furthermore, car- closely linked.In thehumansystem theIL-1receptor riageofatleastonecopyoftheA2allelewasfoundto genes map in chromosome 2 close to the IL-1 gene be associated with reduced bone loss at the spine in cluster, whereas this is not the case in the mouse. early postmenopausal patients (Keen et al., 1998). In normal human subjects, IL-1Ra allele 2 has a clear influence on IL-1Ra circulating levels, i.e. its carrier individualshad10-foldhigherlevels(745ng/mL)than Regulatory sites and corresponding the noncarrier individuals (627pg/mL). This also transcription factors required the presence of the IL-1(cid:12) -511 allele 2 or absence of the IL-1(cid:12) +3953 allele 2, indicating that the IL-1(cid:12) gene participates in the regulation of IL- sIL-1Ra, icIL-1RaI, and icIL-1RaII are translates 1Ra production in vivo (Hurme and Santtila, 1998). originating from the same gene (Figure 1). sIL-1Ra Reciprocally, the presence of the IL-1Ra allele 2 is and icIL-1RaI are produced by the alternative spli- associated with enhanced IL-1(cid:12) production in cing of two different exons 1: 1 and 1 . Exon 1 is s ic ic mononuclear cells in vitro (Santtila et al., 1998). located upstream of exon 1 . Exon 1 splices into s ic The activity of sIL-1Ra promoter is highly exon 1 excluding a large proportion of the leader s controlled by tissue restricted factor(s), being peptide sequence (Butcher et al., 1994). The tran- inactivated in T cells (Jurkat), B cells (Raji), and scription of icIL-1Ra and sIL-1Ra is controlled by epithelial cells (HeLa, HT29) (Smith et al., 1992; two different promoters, P and P , respectively. P ic s ic Butcher et al., 1994). Furthermore, the sIL-1Ra gene is located upstream of exon 1 and P is located ic s is not constitutively expressed in resting or unstimu- upstream of exon 1, in the 9.4kb intron which s lated monocytes, and established myeloid cell lines separates exon 1 from 1 . Three transcription start ic s such as THP1, in contrast to macrophages.Promoter siteshavebeenidentifiedforhumansIL-1RamRNA, elements controllingthe induction ofhumansIL-1Ra Figure 1 IL-1Ra isoforms are generated from the same gene. icIL-1Ra is producedbysplicingofexon1 intoexon1 excludingalargeproportionof ic s the leader peptide sequence (dashed zone). The production of sIL-1Ra is controlled by promoter P. sIL-1Ra pro-peptide is processed and glycosy- s lated prior to secretion. Pic 1ic Ps 1s 2 3 4 Gene 5'UTR 5'UTR 3'UTR Primary RNA 5'UTR 3'UTR mRNA 5'UTR 3'UTR 5'UTR 3'UTR IcIL-1Ra pro-sIL-1Ra Intracellular protein sIL-1RA Secreted protein 322 Danielle Burger and Jean-Michel Dayer Table 1 Association of two allele repeat (IL1RN*2) with inflammatory diseases References Associated diseases Alopecia aerata Cork et al. (1995), Tarlow et al. (1994) Systemic lupus erythematosus Suzuki et al. (1997), Blakemore et al. (1994) Psoriasis Tarlow et al. (1997) Diabetic nephropathy Blakemore et al. (1996) Relapsing/remitting multiple sclerosis de la Concha et al. (1997) Henoch–Scho¨nlein Liu et al. (1997) Sjo¨gren’s syndrome Perrier et al. (1998) Nonassociated diseases Graves’ disease Muhlberg et al. (1998), Cuddihy and Bahn, (1996) Ulcerative colitis Hacker et al. (1997), Bioque et al. (1996) Rheumatoid arthritis Tarlow et al. (1994) Crohn’s disease Mansfield et al. (1994) were identified in mouse (RAW 264.7) and human transcriptionalactivationofsIL-1Ragene(Ohmoriet (THP1)myeloid cell lines(Smith et al., 1992;Butcher al., 1996). etal.,1994).sIL-1Rapromoterisactiveintransfected Intransfectionstudies,apromoter/luciferasefusion cell lines, suggesting a dysregulation in such systems. construct containing 4.5kb of the 50 flanking Nucleotide deletion downstream of -294 had no sequence of the human icIL-1Ra gene exhibited a significant effect on promoter activity, although this pattern of expression in epithelial cell lines and region might contain elements required for the macrophage cell lines similar to that of the inducible activation of sIL-1Ra transcription in endogenous icIL-1Ra gene (Jenkins et al., 1997). myeloid cells. Indeed, most proximal 294 bp of the Mutants of the latter promoter construct indicated human sIL-1Ra promoter are sufficient for full basal that the constitutive expression in epithelial cells was activity and LPS responsiveness. Four LPS-respon- under the control of three positively acting regions sive elements (LRE) have been identified in the sIL- locatedbetweenbases-4525to-1438,-288to-56,and 1Ra promoter. One LRE masking the response to -156 to -49. In contrast, basal activity of the icIL- LPS is located between -294 and -250 (INH). Three 1RaIpromoterintransfectedbutunstimulatedRAW other positive-acting sites were identified between - 264.7cellswasunderthecontrolofaweakinhibitory 250 and -200 (LRE3), -200 and -148 (LRE2), and -93 region located between -4525 and -1438bp and a and -84 (LRE1) (Smith et al., 1994). LRE1, LRE2, strong positive element between -156 and -49. and LRE3 exhibited cooperativity in mediating the Induction of icIL-1Ra transcription by LPS in responses to LPS. LRE1 was identified as an NF(cid:20)B- RAW 264.7 cells was regulated by strong positively binding site. The transcription factors binding to acting DNA regions between bases -1438 to -909 and LRE2 and LRE3 remain to be identified. Recently, -156 to -49. In summary, the proximal region of the two PU.1-binding sites were identified (Smith et al., icIL-1Ra promoter, between bases -156 and -49, 1998), one of which centered at -230 whose mutation containspositivecis-actingelementsthatarerequired resulted in a 50% decrease in LPS-responsive for expression in both epithelial cell and monocytic promoter activity. The other PU.1-binding site, cell lines. which partially overlapped the NF(cid:20)B site, was revealed to be a novel composite NF(cid:20)B/PU.1/GA- binding protein-binding site. Both PU.1-binding sites Cells and tissues that express are major responsive elements for LPS-induced sIL- the gene 1Ra gene expression. Two STAT-binding elements (SBEs) were identified in a region between -250 and - 200.UponIL-4activationofmonocyte-macrophages, sIL-1Raisaninduciblegeneinmostcellswithmaybe STAT6 binds to the more distal SBE, inducing the exception of astrocytes (Liu et al., 1998), whereas IL-1Ra 323 icIL-1Ra is expressed constitutively in keratinocytes Figure2 ThebindingofIL-1RatoIL-1RtypeI and intestinal epithelial cells (for review see Arend does not trigger the formation of a trimolecular et al., 1998). Furthermore, icIL-1RaI mRNA is con- complex, thus blocking signal transduction. stitutively expressed in the epithelial cell lines A431 Residue 145 confers the agonistic (aspartic acid, D) or the antagonistic (lysine, K) activity to the andHT29,butnotinthemacrophagecelllinesRAW ligand. 264.7 and U937, or in the lymphocyte cell lines Raji and Jurkat. However, icIL-1Ra mRNA expression was induced in response to stimulation with LPS in D IL-1b RAW 264.7 cells and to PMA and LPS in U937 cells K (Jenkins et al., 1997). IL-1Ra IL-1RI In normal mice and those injected with LPS, icIL- 1RamRNAwasfoundonlyintheskin,whereassIL- 1RamRNAwasnotdetectedinanytissuesofnormal micebutwasupregulatedinspleen, lung,andliverof LPS-treated mice (Gabay et al., 1997a). In normal rabbits, IL-1Ra was constitutively produced in all IL-1R AcP tissues examined (Apostolopoulos et al., 1996; K D Matsukawa et al., 1997). All tissues produced sIL- 1Ra whereas thymus, cecum, skin, and kidney produced both sIL-1Ra and icIL-1RaI. Low affinity High affinity bimolecular complex trimolecular complex PROTEIN icIL-1RaI, and icIL-1RaII, were also identified in Accession numbers mouse and rabbit (Goto et al., 1992; Cominelli et al., 1994; Zahedi et al., 1994; Gabay et al., 1997a; SwissProt: Matsukawa et al., 1997). Human: p18510 Rabbit: p26890 Description of protein Mouse: p25085 Rat: p25086 cDNAsequencesforbovineandhorseIL-1Rahave IL-1Ra displays a hydrophobic core and the typical been described recently (Kato et al., 1997; Kirisawa (cid:12)-trefoilstructurewith12(cid:12) strandswhichisfoundin et al., 1998; Howard et al., 1998) but are not yet the IL-1 family of proteins (Stockman et al., 1992, accessible in data banks. 1994; Vigers et al., 1994). Residues Trp16, Gln20, Tyr34, Gln36, and Tyr147 are crucial for the binding of sIL-1Ra to the receptor (Evans et al., 1995). Sequence Residue 145(Lys in sIL-1Ra, Asp in IL-1(cid:12)) is crucial indeterminingagonistorantagonistactivity(Juetal., Human sIL-1Ra precursor is a 177 amino acid 1991). However, this region does not interact directly glycoprotein comprising a signal sequence (amino with the receptor, either in IL-1Ra or in IL-1(cid:12) acids 1–25), a disulfide bound between Cys69 and (Schreuder et al., 1997), suggesting that this residue Cys116 (Schreuder et al., 1995), a potential N- might be important for the interaction with the IL-1 glycosylation site at Gln109 and a cis-Pro at position receptor accessoryprotein(IL-1RAcP) (Arendet al., 53. Human 18kDa icIL-1Ra displays a sequence 1990; Schreuder et al., 1997) (Figure 2). similar to pro-sIL-1Ra with the N-terminal sequence MAL instead of MEICRGLRSHLITLLLFLFHS, Discussion of crystal structure givingrisetoa159aminoacidunglycosylatedprotein (Haskillet al., 1991).Rabbit,rat, mouse,bovine, and horsesIL-1Rapredictedaminoacidsequencesdisplay The 3D structure of IL-1Ra has been determined by around 75% sequence homology with human sIL- X-ray crystallography and in solution by NMR 1Ra, the rat and mouse sIL-1Ra precursor exhibiting spectroscopy (Stockman et al., 1992, 1994; Vigers asignalsequenceof26residues,beingoneaminoacid et al., 1994; Schreuder et al., 1995, 1997). IL-1Ra has longer than that of human and rabbit sIL-1Ra. The 12 (cid:12) strands connected by loops. Six of the (cid:12) strands threeformsofIL-1Rafoundinhumans,i.e.sIL-1Ra, formabarrelstructurethatisclosedatoneendbythe 324 Danielle Burger and Jean-Michel Dayer remaining (cid:12) sheet. The overall structure of IL-1Ra is icIL-1Raorbothforms.However,byfarthemajority similar to that of IL-1(cid:12) and IL-1(cid:11). Differences in of the studies have focused on the expression of sIL- the amino acid sequence of the two types of protein 1Ra in monocyte-macrophages, neutrophils, and account for the interactions with the receptor, fibroblasts and of icIL-1Ra in epithelial cells. The triggering or not signal transduction through the induction of IL-1Ra production in various cells has formation of the trimolecular complex IL-1/IL-1RI/ been extensively reviewed (Lennard, 1995; Dinarello, IL-1R AcP. 1996; Arend et al., 1998). The differential expression of IL-1 and IL-1Ra in different cell types of the Important homologies central nervous system has recently been described. IL-1Ra (protein and mRNA) is constitutively expressed in neurons of the paraventricular nucleus Recombinant human sIL-1Ra binds to mouse, and supraoptic nucleus. After administration of bovine, and rabbit IL-1 receptor with similar avidity. endotoxin in rats, no additional cell expressed This suggests that IL-1Ra has been well conserved immunoreactive IL-1Ra whereas IL-1(cid:11) and IL-1(cid:12) through evolution. Indeed, the amino acid sequences were induced in the same cell types, i.e. macrophages of all the mammal IL-1Ra described to date display inmeningesandchoroidplexusandmicroglialcellsin >75% homology, e.g. the deduced amino acid various brain regions. This suggests that in the brain sequence of bovine IL-1Ra demonstrated 80%, IL-1Ra is constitutively expressed by cell types other 78%, 78%, 77%, and 76% homology with human, than those expressing IL-1 (van Dam et al., 1998). mouse, rat, rabbit, and horse sequences, respectively (Kirisawa et al., 1998). There is poor sequence homology between IL-1Ra and IL-1(cid:11) or IL-1(cid:12), Eliciting and inhibitory stimuli, although all three proteins are derived from a unique including exogenous and primordial precursor gene which gave rise to a commonIL-1(cid:11)/IL-1(cid:12) geneandtheIL-1Rageneafter endogenous modulators a gene duplication event which occurred some 350 million years ago (Eisenberg et al., 1991). The production of IL-1Ra, and particularly that of sIL-1Ra, has been extensively studied. Indeed, since Posttranslational modifications sIL-1Ra is usually produced by the same cell as IL-1, the identification of factors able to modulate TheonlyposttranslationaleventwhichoccursinsIL- differentially these productions has remained a 1Rais glycosylation.Indeed,apotential N-glycosyla- challenge. LPS induces both IL-1(cid:12) and IL-1Ra in tion site is present in the IL-1Ra sequence of all monocytes. However, the expression kinetics of the species described to date. However, the presence of two proteins are different, the production of IL-1(cid:12) theoligosaccharidemoietydoesnotseemtoaffectthe andsIL-1Rareachingitspeakat2hoursand4hours, inhibitory activity of sIL-1Ra since both natural, respectively (Vannier et al., 1992). Similarly, differ- glycosylated sIL-1Ra and recombinant, unglycosyl- entialproductionofIL-1(cid:12) andsIL-1Rawasobserved ated sIL-1Ra display similar avidity to bind to the in vivo. LPS administered to human volunteers IL-1 receptor. Therefore, glycosylation might protect induced the production of IL-1(cid:12) and sIL-1Ra which sIL-1Ra from proteolytic degradation and thus peaked 1 hours and 2 hours after injection, prolong its lifespan in the extracellular space. Since respectively. Furthermore, peak plasma concentra- theintracellularformsofIL-1Raarenottranslatedin tions of IL-1Ra were about 100-fold higher than the endoplasmic reticulum and are consequently not those of IL-1(cid:12) (Granowitz et al., 1991). In vitro, the processed through the Golgi, icIL-1RaI and icIL- most potent inducer of IL-1Ra in monocytes is 1RaII are not glycosylated. adherent IgG which weakly induces IL-1(cid:12) produc- tion. In monocytes activated by adherent IgG, high concentrations of LPS reduce the expression of sIL- CELLULAR SOURCES AND 1Ra but not that of IL-1(cid:12). Together, these studies indicatethatdifferentsignalingpathwaysareinvolved TISSUE EXPRESSION in the induction of IL-1(cid:12) and IL-1Ra production. Another important inducer of sIL-1Ra in monocytes Cellular sources that produce is direct cellular contact with stimulated T cells (Burger and Dayer, 1998) which also concomitantly It is very likely that all cell types able to produce induces the production of IL-1 and sIL-1Ra. In the IL-1(cid:11) and/or IL-1(cid:12) will also express sIL-1Ra or latter system, IL-1(cid:12) and sIL-1Ra production is IL-1Ra 325 controlled by different intracellular mechanisms IL-1(cid:12) in monocytic cells. Activin A inhibits the involving the differential effect of serine/threonine production of IL-1(cid:12) and enhances that of IL-1Ra in phosphatases (Vey et al., 1997), the latter enzymes activated THP1 and U937 human monocytic cells. It having a positive effect on the induction of sIL-1Ra is noteworthy that activin A regulates cytokine and a negative one on the induction of IL-1(cid:12). This production at a posttranscriptional level (Ohguchi et has been confirmed in a study demonstrating that al., 1998). IFN(cid:12) also differentially modulates IL-1Ra Raf-1kinase, aserine/threoninekinase, is involvedin and IL-1(cid:12) production in PHA-stimulated peripheral one of the pathways leading to sIL-1Ra production. blood mononuclear cells (PBMCs) by inhibiting the Interestingly, the Raf-1 and LPS pathways are production of IL-1(cid:12) and enhancing that of IL-1Ra mutually antagonistic (Guthridge et al., 1997). The (Coclet-Nininetal.,1997).Retinoicaciddifferentially ability of stimulated T lymphocytes to induce IL-1 regulates IL-1(cid:12) and IL-1Ra production by alveolar and IL-1Ra is differentially modulated by drugs such macrophages (Hashimoto et al., 1998). as leflunomide (De´age et al., 1998). IL-3 induces sIL- 1Ra production in monocytes to levels equivalent to RECEPTOR UTILIZATION those induced by GM-CSF and LPS, whereas IL-1(cid:11) and IL-4 are weak inducers (Jenkins and Arend, IL-1Ra secreted form interacts with both IL-1 1993). receptors. Natural and recombinant (unglycosylated) Macrophages differentiated either in vivo (lung sIL-1Ra display similar affinity for the mouse IL-1 alveolarmacrophages,synoviocytes)or in vitro(GM- receptor I to that of human IL-1(cid:11) and IL-1(cid:12) CSF-treatedmonocytes)spontaneouslyproducelarge (dissociation constant(cid:129)200nM) (Seckinger et al., amounts of sIL-1Ra (Roux-Lombard, 1998). This 1987; Hannum et al., 1990; Eisenberg et al., 1990). production is increased by GM-CSF and IL-4 sIL-1Ra binding to IL-1 receptor I does not induce whereas LPS remains ineffective. Furthermore, signaling because it does not engage IL-1R AcP C-reactive protein (CRP) triggers the production of (Cullinan et al., 1998). However, in addition to sIL- IL-1 and sIL-1Ra in peripheral blood monocytes but 1Ra,thebiologicalactivityofIL-1iscounterbalanced inhibits it in alveolar macrophages (Tilg et al., 1993; by IL-1 soluble receptors (IL-1sR) which bind IL-1 Pue et al., 1996). and diminish the free concentration of soluble In polymorphonuclear neutrophils (PMNs), sIL- cytokine, thus hampering its binding to the cell 1Ra, but not icIL-1Ra, is mainly induced by LPS, surface receptor. Since IL-1sR can also bind IL-1Ra, GM-CSF, and TNF(cid:11) (Malyak et al., 1994). the two types of inhibitors might abolish their Interestingly, TNF(cid:11) is a poor inducer of sIL-1Ra in respective inhibitory activity. Interestingly, the inhib- monocytes. The induction of sIL-1Ra in PMNs is itory effect of human sIL-1Ra is enhanced by type II synergistically enhanced by IL-4 and IL-10, but not IL-1sR and hindered by type I IL-1sR (Burger et al., by IL-13 and TGF(cid:12). IFN(cid:13) increases the release of 1995). sIL-1Ra in LPS- and fMLP-stimulated PMNs Recombinant human sIL-1Ra, icIL-1RaI, and (McDonald et al., 1998). icIL-1RaII display similar affinity constants for type In human microglia, IFN(cid:12) induces IL-1Ra II IL-1sR (2.9–3.8(cid:2)107M(cid:255)1), whereas icIL-1RaII production and enhances LPS- and IL-4-induced binds type I IL-1sR with a 4- to 5-fold lower affinity IL-1Ra production, whereas it suppresses LPS- and constant(3.0(cid:2)109M(cid:255)1)thansIL-1RaandicIL-1RaI IL-1-induced IL-1(cid:12) production. In contrast, IFN(cid:13) (1.2and1.4(cid:2)1010M(cid:255)1)(Malyaketal.,1998).Inspite inhibits both IL-1(cid:12)- and IL-1Ra-induced production of very similar affinity constants for IL-1 receptor (Liu et al., 1998). sIL-1Ra and IL-1 production by typeI,largemolarexcess(10-to100-fold)ofsIL-1Ra PMNs is also differentially induced by microcrystals, is required to efficiently inhibit IL-1 activity in vitro favoring that of IL-1 (Roberge et al., 1994). (Arend et al., 1990). Similarly, preinjection of 100- The potent induction of sIL-1Ra by either LPS or to 1000-fold molar excess of sIL-1Ra is required to GM-CSF might be modulated by other cytokines block systemic responses to IL-1 in animals (Fischer such as IL-4, IL-10, and IL-13 which potentiate sIL- et al., 1991). 1Ra production and simultaneously inhibit IL-1(cid:12) production (Jenkins and Arend, 1993; Jenkins et al., IN VITRO ACTIVITIES 1994). Interestingly, although exogenous IL-1 is a poor sIL-1Ra inducer in monocytes, the induction of endogenous IL-1 by TGF(cid:12) in monocytes is required In vitro, sIL-1Ra has mainly been used as a specific forthetriggeringofsIL-1Raproduction(Wahlet al., inhibitor of IL-1(cid:11) and IL-1(cid:12) activity, since it is a 1993). Activin A, a member of the TGF(cid:12) family, has cytokine whose only known action is the competitive beenshowntoregulatetheproductionofIL-1Raand inhibition of the binding of IL-1 to its receptor. 326 Danielle Burger and Jean-Michel Dayer In vitro findings IL-1Ra or overproduce it under the control of the endogenous promoter. IL-1Ra has been shown to inhibit all described IL-1 activities due to the binding of the agonist to its Knockout mouse phenotypes receptor, although the concentration required to abolish IL-1 effects is usually a 10- to 100-fold molar Mice lacking IL-1Ra are more susceptible than excess of IL-1Ra over IL-1. More studies are controls to lethal endotoxemia but less susceptible to required todefinethe functions ofintracellular forms listeriosis (Hirsch et al., 1996). Furthermore, IL-1Ra of IL-1Ra. knockout mice had a significantly earlier onset of collagen-inducedarthritis, with increased severity (Ma Regulatory molecules: Inhibitors et al., 1998). This confirms the important role of IL- 1RaintheregulationofIL-1activityduringinfection and enhancers and inflammation. More intriguing is the fact that IL-1Ra knockout mice have decreased body mass Since it does not induce signal transduction, the compared with wild-type controls or IL-1 knockout ‘activity’ of sIL-1Ra is regulated only by its levels of miceduetogrowthretardationafterweaning(Hirsch production which are controlled by the above et al., 1996; Horai et al., 1998). This implies that mentioned inducers, inhibitors, and enhancers. IL-1Ra plays an important but still elusive part in normal physiology. IN VIVO BIOLOGICAL Transgenic overexpression ACTIVITIES OF LIGANDS IN ANIMAL MODELS TheeffectofIL-1Raoverexpressionhasbeenassessed inseveralanimalmodelsofdiseaseinwhichIL-1was Normal physiological roles claimed to be involved. Mice overexpressing the IL- 1Ra gene under the control of its endogenous Inadditiontothe constitutiveexpressionoficIL-1Ra promoter had a significant reduction in both inepithelialcellsandkeratinocytes,sIL-1Raseemsto incidence and severity of collagen-induced arthritis be expressed at low levels in normal human subjects suggesting that the endogenous expression of IL-1Ra and animals, despite some species-dependent differ- isacriticaldeterminantofsusceptibilitytothedisease ences. However, the function of IL-1Ra in normal, (Ma et al., 1998). Interestingly, after immunization i.e. ‘noninflammatory’ conditions, remains elusive. It with type II collagen, IL-1Ra mRNA was over- has been postulated that tissue IL-1Ra is involved in expressedinthespleens,butnotintheinjectionpaws, health maintenance by masking coexisting IL-1 of transgenic mice (Ma et al., 1998). On the other activity in tissues (Matsukawa et al., 1997). In hand, IL-1Ra overproducers are protected from the healthy rabbits, sIL-1Ra is produced constitutively lethal effects of endotoxin while being more in most tissues (lung, liver, spleen, thymus, cecum, susceptible to listeriosis. Following an endotoxin skin, kidney, heart, and brain), while thymus, cecum, challenge serum levels of IL-1 are decreased in skin, and kidney produced both sIL-1Ra and icIL- IL-1Ra-null mice and increased in IL-1Ra overpro- 1Ra. In contrast, IL-1 activity has not been detected ducersincomparisontocontrols(Hirschetal.,1996). in any tissues with the exception of skin and heart, Transgenic mice with distal airway epithelial cell but became detectable after preincubation of the expression of human IL-1Ra were partially protected samples with neutralizing antibodies to IL-1Ra, from IL-1(cid:11)-induced airway inflammation and injury demonstrating that constitutive but low expression (Wilmott et al., 1998). of sIL-1Ra inhibits IL-1 activity in physiological conditions. On the other hand, studies on chick Pharmacological effects embryo fibroblasts have shown that the balance between the secreted cytokine network and IL-1Ra levels may represent a homeostatic mechanism aimed Since IL-1Ra is able to counteract the proinflamma- at controlling normal embryogenesis (Bodo et al., tory effects of IL-1, recombinant sIL-1Ra has been 1998). used as therapeutic agent in many animal disease The physiological roles of endogenously produced models. These models were recently reviewed (Arend IL-1Rahavebeeninvestigatedinmicethateitherlack et al., 1998) and covered mainly infectious and IL-1Ra 327 inflammatory diseases. Exogenous IL-1Ra, usually IL-1-dependent mechanism. Therefore, coinfusion of administered intravenously, led to at least partial glucocorticoids and cytokine should favor the pro- prevention or treatment of disease. However, the duction of IL-1Ra over that of IL-1(cid:12). necessity of systemic IL-1Ra injections in large amounts limits the application of this therapy in human patients. New approaches have recently been PATHOPHYSIOLOGICAL ROLES tested applying exogenous IL-1Ra directly at the IN NORMAL HUMANS AND inflammatory site. These gene therapy trials were mainly carried out in animal models of rheumatoid DISEASE STATES AND arthritis and osteoarthritis using either adenoviral DIAGNOSTIC UTILITY gene transfer into the joint or ex vivo gene transfer in synoviocytes and autograft (Arend et al., 1998). The role of IL-1Ra in normal physiology of healthy Interestingly, endogenously produced IL-1Ra was humans remains elusive. Only animal models have mainly chondroprotective,displaying onlyweak anti- shed light on the possible role of IL-1Ra in normal inflammatory properties. Therefore, IL-1Ra gene biological processes. Indeed, in contrast with IL-1 therapy displays some efficacy in inflamed animal knockout mice, IL-1Ra knockout mice display joints mainly by diminishing cartilage destruction. growth retardation, suggesting an important role for Few trials have been carried out in other systems, IL-1Ra in normal physiology where it could mask although the transfer of IL-1Ra gene into neuronal coexisting IL-1 activity in tissues. However, addi- cells, hematopoietic stem cells and lungs has been tional studies are required to clarify the roles of successful (Davidson et al., 1993; Boggs et al., 1995; IL-1Raisoformsinnormalbiologicprocessesparticu- McCoyetal.,1995).IL-1Raoverexpressioninmouse larly in humans. Because sIL-1Ra levels are elevated brain by adenoviral vector attenuates the activation in the peripheral blood of patients with various of inflammatory cells during focal cerebral ischemia diseases, ithasrecentlybeensuggestedthatIL-1Rais (Yang et al., 1998). IL-1Ra may also be upregulated an acute phase protein (Gabay et al., 1997b). in rat brain by peripheral administration of kainic Furthermore, the use of neutralizing antibodies has acid (Eriksson et al., 1998). demonstrated the importance of IL-1Ra as a natural anti-inflammatory protein in animal models of diseases. In analogy a similar role for IL-1Ra has Interactions with cytokine network been suggested in humans. Indeed, as recently reviewed (Arend et al., 1998), extensive evidence IL-1 being a potent inducer of many other cytokines, indicates that IL-1Ra is produced by many tissues in the inhibition of its effects by IL-1Ra clearly has an healthy humans and that its production is upregu- important impact on the cytokine network. Further- lated in the host response to infection and acute and more, since IL-1 may induce the production of its chronic inflammation. inhibitor, sIL-1Ra might regulate its own production in some circumstances. Normal levels and effects Endogenous inhibitors and Except for allele 2 carriers, low levels of circulating enhancers sIL-1Ra have been detected in healthy humans (400– 800pg/mL) which were drastically increased in IL-1Ra is usually produced by the same cells which various pathological conditions (Table 2). The produce IL-1(cid:12) and/or IL-1(cid:11). Besides, similar stimuli increase in IL-1Ra level in such diseases does not inducetheproductionofbothagonistandantagonist seemtocounteractefficientlytheinflammatoryeffects cytokines.Therefore,manystudieshavebeendevoted of IL-1. However, studies using neutralizing anti-IL- to the study of factors affecting the balance between 1Ra antibodies in animal models of disease indicate IL-1RaandIL-1production,i.e.IL-4,IL-13,andIL- an anti-inflammatory effect of endogenous IL-1Ra. 10 favoring the production of sIL-1Ra over that of Ontheotherhand,thelevelsofIL-1Rawerefoundto IL-1(cid:12). Interestingly, both TH1 and TH2 cytokines, be locally enhanced in acute and chronic disorders i.e. IFN(cid:13) and IL-4, antagonize the inhibition of such as rheumatoid arthritis (Firestein et al., 1992; sIL-1Ra production by LPS-stimulated monocytes Malyaketal.,1993),osteoarthritis,andLymearthritis in the presence of glucocorticoids (Parsons et al., (Miller et al., 1993), and nonperforating Crohn’s 1997), although IFN(cid:13) but not IL-4 acts through an disease (Gilberts et al., 1994). 328 Danielle Burger and Jean-Michel Dayer Table 2 Diseases with increased circulating IL-1Ra level Disease Reference Chronic arthritis Jouvenne et al. (1998) Insulin-dependent diabetes mellitus Netea et al. (1997) Systemic lupus erythematosus Sturfelt et al. (1997) Gynecological cancers Fujiwaki et al. (1997) Pediatric sepsis syndrome Samson et al. (1997) Pre-eclampsia Kimya et al. (1997) Acute myocardial infarction Shibata et al. (1997) Graft-versus-host disease Schwaighofer et al. (1997) Hemophagocytic lymphohistiocytosis Henter et al. (1996) Patients with burns Endo et al. (1996) Bronchial asthma Yoshida et al. (1996) Chronic renal failure Deschamps-Latscha et al. (1995) Systemic chronic juvenile arthritis De Benedetti et al. (1995) Polymyositis/dermatomyositis Gabay et al. (1994) Septic shock Rogy et al. (1994) Role in experiments of nature as prognostic markers or for therapeutical usage (Jouvenne et al., 1998). In renal transplant recipients, and disease states patients with high IL-1Ra/IL-1(cid:12) ratios in urine are less prone to acute allograft rejection than patients The serum levels of sIL-1Ra are elevated in many with low IL-1Ra/IL-1(cid:12) ratios (Teppo et al., 1998). pathologies. However, only few studies contemplate Rejection episodes in orthotopic heart transplanta- the determination of IL-1Ra as a diagnostic test. In tion are accompanied by a renewed and more inflammatory bowel disease the balance between IL-1 pronouncedelevationinIL-1Ra(serumlevelsbeyond and sIL-1Ra might influence disease expression. It 4000pg/mL) for at least 2 days, indicating that IL- was demonstrated that patients with active Crohn’s 1Ra might have a predictive value in the detection disease had significantly higher serum sIL-1Ra levels of acute allograft rejection (Thiele et al., 1998). than patients with active ulcerative colitis. Therefore, levels of sIL-1Ra in the peripheral blood of patients with inflammatory bowel disease are of clinical IL-1RA IN THERAPY relevance, representing a marker of disease activity and a possible differential diagnostic marker (Propst Preclinical – How does it affect et al., 1995). More recently, it has been shown that disease models in animals? serum levels of sIL-1Ra and IL-6 were enhanced 2 days before the clinical manifestation of neonatal sepsis (Kuster et al.,1998).While circulatingsIL-1Ra As stated above, the therapeutic use of sIL-1Ra has levels are increased in patients at risk for acute been tested in many animal models of diseases respiratorydistresssyndrome(ARDS)whodie,itdoes (Table 3). This has been extensively reviewed (Arend not predict the development of the syndrome et al., 1998). Some studies published in 1998 have (Parsons et al., 1997). Furthermore, it was recently shownthatIL-1Rawasalsoefficientinthetreatment suggested that in chronic polyarthritis elevated levels of experimental allergic encephalomyelitis (EAE), the of IL-1Ra may reflect increased production and animal model for multiple sclerosis (Badovinac et al., activity of IL-1, in contrast with the levels of 1998), although IL-1Ra was administered before endogenous type II IL-1sR which may constitute a disease onset. Indeed, dark Agouti rats which were natural anti-inflammatory factor. This should be treatedduringtheinductionphaseofEAE(days0–6) taken into account when considering these molecules with IL-1Ra (350(cid:22)g/rat/day) developed milder