Interleukin-1 receptor antagonist competitively inhibits the binding of interleukin-1 to the type II interleukin-1 receptor.

The interleukin-1 receptor antagonist (IL-1ra) inhibits the binding of interleukin-1 (IL-1) to T-cell lines possessing the type I IL-1 receptor; evidence has been published (Carter, D. B., Deibel, M. R. J., Dunn, C. J., Tomich, C. S., Laborde, A. L., Slightom, J. L., Berger, A. E., Bienkowski, M. J., Sun, F. F., McEwan, R. N., Harris, P. K. W., Yem, A. W., Waszak, G. A., Chosay, J. G., Sieu, L. C., Hardee, M. M., Zurcher-Neely, H. A., Reardon, I. M., Heinrickson, R. L., Truesdell, S. E., Shelly, J. A., Eessalu, T. E., Taylor, B. M., and Tracey, D. E. (1990) Nature 344, 633-638; Hannum, C. H., Wilcox, C. J., Arend, W. P., Joslin, F. G., Dripps, D. J., Heimdal, P. L., Armes, L. G., Sommer, A., Eisenberg, S. P., and Thompson, R. C. (1990) Nature 343, 336-340) that IL-Ira does not bind to the type II IL-1 receptor (IL-1RtII). In this study we examined the ability of human recombinant IL-1ra to block the binding of IL-1 to the IL-1RtII on human polymorphonuclear leukocytes (PMN) and Raji human B-lymphoma cells. The binding of 125I-IL-1 beta to PMN was competively inhibited by IL-1ra. IL-1 beta was more potent in inhibiting the binding of 125I-IL-1 beta than IL-1ra. Incubating PMN with 125I-IL-1ra in the presence of increasing concentrations of IL-1 beta or IL-1ra showed that IL-1 beta was an approximately 40-fold more potent inhibitor of binding of 125I-IL-1ra than unlabeled IL-1ra. The IL-1ra was approximately 500-fold less potent in inhibiting the binding of 125I-IL-1 alpha than IL-1 alpha. IL-1ra was also able to competitively inhibit binding of 125I-IL-1 beta to Raji cells. PMN or Raji cells were also incubated with 125I-IL-1 in the absence or presence of IL-1 or IL-1ra. After cross-linking of IL-1 to cells followed by specific immunoprecipitation, sodium dodecyl sulfate-polyacrylamide gel electrophoresis showed a band at 85 kDa corresponding to the 68-kDa IL-1RtII. However, in the presence of an excess of either unlabeled IL-1 or IL-1ra, the 85-kDa IL-1.IL-1RtII complex was not present. These findings demonstrate that the IL-1ra recognizes and blocks IL-1 binding to the IL-1RtII.

E., Shelly, J. A., Eessalu, T. E., Taylor, B. M., and Tracey, D. E. (1990) Nature 344,633-638; Hannum, C. H., Wilcox, C. J., Arend, W. P., Joslin, F. G., Dripps, D. J., Heimdal, P. L., Armes, L. G., Sommer, A., Eisenberg, S. P., and Thompson, R. C. (1990) Nature 343, 336-340) that IL-Ira does not bind to the type I1 IL-1 receptor . In this study we examined the ability of human recombinant IL-lra to block the binding of IL-1 to the IL-1RtII on human polymorphonuclear leukocytes (PMN) and Raji human B-lymphoma cells. The binding of 12sI-IL-l/3 to PMN was competitively inhibited by IL-lra. IL-18 was more potent in inhibiting the binding of '261-IL-18 than IL-Ira. Incubating PMN with '251-IL-lra in the presence of increasing concentrations of IL-18 or IL-I r a showed that IL-18 was an -40-fold more potent inhibitor of binding of '2sI-IL-lra than unlabeled IL-lra. The IL-lra was -500-fold less potent in inhibiting the binding of ' "1-IL-la than IL-la. IL-lra was also able to competitively inhibit binding of 12sI-IL-l, to Raji cells. PMN or Raji cells were also incubated with '251-IL-l in the absence or presence of IL-1 or IL-lra. After cross-linking of IL-1 to cells followed by specific immunoprecipitation, sodium dodecyl sulfate-polyacrylamide gel electrophoresis showed a band at 85 kDa corresponding to the 68-kDa IL-1RtII. However, in the presence of an excess of either unlabeled IL-1 or IL-lra, the 85-kDa IL-1 eIL-lRtII complex was not present. These findings demonstrate that the IL-lra recognizes and blocks IL-1 binding to the IL-1RtII.
* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U. Interleukin-1 (IL-1)' is the term used to describe two 17.5-kDa polypeptides, IL-la and IL-lp, which share a wide spectrum of inflammatory and immunological properties (3). ILla and IL-lp possess 25% amino acid similarity and bind to the same receptors. However, IL-la binds preferentially to the 80-kDa type I IL-1 receptor (IL-1RtI) found on T-cells (4)(5)(6)(7), fibroblasts (4), and keratinocytes (8); IL-lp binds best to the 68-kDa type I1 IL-1 receptor (IL-1RtII) on B-cells (6,7,9) andpolymorphonuclear leukocytes (PMN) (10,ll). This differential binding of IL-la and IL-1p is observed even though there is 28% amino acid identity between the extracellular portions of the IL-1RtI and the IL-1RtII (12). The cytoplasmic portion of the IL-1RtII is truncated in comparison with the IL-1RtI.
Binding of both IL-la and IL-lp to the IL-1RtI on T-cells can be inhibited by the IL-1 receptor antagonist (IL-lra), an 18-kDa protein which has 19% amino acid similarity with ILla and 26% amino acid similarity with IL-la (13). Evidence has also been published that IL-lra does not bind to the IL-lRtII (1, 2). Using competitive binding and cross-linking studies, we demonstrate that IL-lra inhibits the binding of both IL-la and IL-lp to the IL-1RtII on polymorphonuclear leukocytes and Raji human B-lymphoma cells.

Materials
Recombinant human IL-la was a kind gift of Dr. Peter Lomedico (Hoffmann-LaRoche). Recombinant human IL-la was donated by Dr. Aldo Tagliabue (Sclavo Research Centre, Siena, Italy). Recombinant human IL-lra was provided by Dr. Robert C. Thompson (Synergen Inc., Boulder, CO). Radiolabeling of IL-16 and IL-lra with ' ' ' I (Du Pont-New England Nuclear) was performed using the tecbnique of Bolton and Hunter. IL-la was radiolabeled with ' ' ' I (Du Pont-New England Nuclear) using chloramine T (Sigma).

Methods
Human PMN-PMN were prepared from the blood of healthy human volunteers. Blood was drawn into syringes containing heparin (20 units/ml final concentration, LyphoMed Inc., Rosemont, IL). Blood was separated by centrifugation through Ficoll (Sigma) and Hypaque (90%, Winthrop Laboratories, New York). Following aspiration of plasma and mononuclear cells, the PMN were recombined with plasma (1 m1/5-ml red cell mass) in a 250-ml graduated cylinder. An equal volume of 3% dextran (M, 266,000, Sigma) in sterile 0.9% sodium chloride (Abbott) was added and the graduated cylinder

IL-lra Inhibits Binding
of IL-1 to  inverted three times. One hour later the supernatant was removed and centrifuged at 250 X g for 10 min at 4 "C. Pellets underwent hypotonic lysis at 4 "C with 0.2% sodium chloride followed 45 s later by an equal volume of 1.6% sodium chloride. Following centrifugation at 250 X g for 10 min at 4 "C, hypotonic lysis was repeated. As determined by Diff-Quik (American Scientific Products, Aguada, Puerto Rico) staining, the cell population was 97-99% PMN with >99% of the cells viable by trypan blue exclusion.
Cell Culture-Raji cells were obtained from American Type Culture Collection. EL4-6.1 cells were a kind gift of Dr. Robert Newton (E. I. Du Pont de Nemours & Co., Glenolden, PA). Both cell lines were cultured in RPMI 1640 containing 5% fetal calf serum.
Receptor Binding Assays-Cells were washed twice in cold RPMI and resuspended at different concentrations in binding buffer. Aliquots of cells were rotated in duplicate at 10 rpm at 4 "C overnight with radiolabeled ligands and different concentrations of either unlabeled IL-la, IL-10, or IL-lra. Total volume of cell suspension was 150 pl. Cell-bound radioactivity was separated from free radioactivity by centrifugation through oil (Silicones, General Electric, Waterford, NY) at 14,000 X g for 90 s at 4 "C. Tips of the tubes containing the cell pellet were cut and counted. Radioactivity was measured using a r-scintillation counter.
Cross-linking of 1251L-lD to IL-IR-Following two washes in cold RPMI, cells were resuspended in binding buffer. Either 2 X 10' PMN, 3 X lo7 Raji cells, or 4 X 10' EL4 cells were rotated at 10 rpm at 4 "C overnight with 1 nM 9 -I L -l P and either 1 p M IL-10 or 10 p M IL-Ira. Similar numbers of PMN, Raji cells, or EL4 cells were incubated under identical conditions with 1 nM '251-IL-la and either 1 M IL-la or 100 p M IL-lra. Total volume of cell suspension was 1000 p1. Cells were then washed with cold PBS and cross-linked using 2.7 mM bis(sulfosuccinimidy1)suberate (BS,, Pierce Chemical Co.) for 2 h at 4 "C. Following a wash with cold PBS the cells were resuspended in CLB. After 30 min on ice, cell debris was removed by centrifugation at 13,000 X g for 15 min at 4 "C.
Immunoprecipitation-Using the methods of Rangnekar et al. (14) and Clark et al. (15), rabbit anti-human IL-10 antibody was used to immobilize the solubilized IL-10. IL-10 complexes. Briefly, 100 p1 of packed IgGSorb (The Enzyme Center, Malden, MA) was incubated with 50 pg of affinity-purified goat anti-rabbit IgG (Southern Biotechnology Associates, Birmingham, AL) at 10 rpm for 6 h at 4 "C. After washing twice in CLB, the conjugated IgGSorb was resuspended in 1000 pl of cold PBS. Aliquots of conjugated IgGSorb were then mixed with ' /IO volume of rabbit anti-human IL-10 and incubated at 10 rpm overnight at 4 "C. The anti-human IL-ID-conjugated IgGSorb was washed twice with CLB before being resuspended in 160 pl of CLB. Prior to immunoprecipation, cell lysates were incubated for 30 min on ice with 1/10 volume of IgGSorb conjugated with goat antirabbit IgG to remove nonspecific binding proteins. PMN, Raji cells, and EL4 cell lysates were combined with ' / a volume of either antihuman IL-l@-conjugated IgGSorb or anti-human IL-la-conjugated IgGSorb and then incubated at 10 rpm overnight at 4 "C. Immunoprecipitates were pelleted by centrifugation at 13,000 X g for 15 min at 4 "C followed by washing in CLB. Supernatants were frozen at SDS-PAGE-The immunoprecipitated IL-lo.IL-1R complexes were resolved by SDS-PAGE with a 10% polyacrylamide gel. Each immunoprecipitate was solubilized in 20 p1 of loading buffer (62 mM Tris-HC1, pH 6.8, 10% w/v glycerol, 1% v/v SDS, and 0.0005% w/v bromphenol blue) and heated at 100 "C for 2 min. Prestained protein molecular weight standards (GIBCO/BRL) were applied to the gel. The gel was dried with heat and vacuum. Autoradiography was performed by exposure of XAR-5 film (Kodak, Rochester, NY) for 10 days at -70 "C using an intensifying screen (Cronex Lightening Plus, Du Pont).

RESULTS
T o determine whether IL-lra is able to interact with the IL-lRtII, PMN were incubated with -450 pM "51-IL-lp in the presence of increasing concentrations of IL-1p or IL-lra. Fig. 1 depicts the dose-dependent inhibition of 1251-IL-lp binding by either IL-1p or IL-lra. By using the Cheng-Prusoff relationship (16) and a KD for IL-1p on PMN of 320 pM (lo), the inhibition constant (KDi) for IL-lra is 7000 pM.
To further investigate the mechanism by which IL-lra inhibits the binding of IL-1 to PMN, increasing numbers of PMN (1O6-1O7/15O pl) were incubated with -350 PM lZ5I-IL-Ira in the absence or presence of 370 nM IL-lra. IL-lra inhibited lZ5I-IL-lra from binding to these PMN (data not shown). Subsequently, PMN were incubated with lZ5I-IL-lra in the presence of increasing concentrations of IL-lp or IL-Ira. As shown in Fig. 2, IL-1p is an -40-fold more potent inhibitor of '251-IL-lra binding to PMN than is IL-lra. This finding is consistent with the data presented in Fig. 1.
As shown in Fig. 3, IL-Ira competitively inhibits 1251-IL-lcu binding to PMN. Using '251-IL-la, the concentration of ILla! which effectively inhibits 50% of the specific radioligand binding (IC50) is -500-fold less than the ICso for IL-lra.
To demonstrate that the aforementioned findings are applicable to other cells having the IL-lRtII, Raji human Blymphoma cells were examined. AS shown in Fig. 4, IL-lra competitively inhibited binding of '251-IL-lp to Raji cells. Using the data presented in Fig. 4 and a KO of 2200 pM for IL-lp on Raji cells (7), the calculated K D i for IL-lra is 7000 P" To confirm that IL-lra recognizes the IL-lRtII, PMN, Raji cells, and EL4 cells were incubated with 1 nM '251-IL-lp in the presence or absence of IL-1p or IL-lra. Subsequently, these cells were cross-linked with BSs and then lysed. The Effect of different Concentrations of IL-la and IL-Ira on binding of '2JI-IL-la to PMN. PMN (2 X IO') were incubated overnight a t 4 "C with -150 PM '2sI-IL-la in the presence of increasing concentrations of IL-la (0) or IL-Ira (0). Nonspecific binding determined in the presence of 37 nM IL-la was 28% of total binding. This experiment represents one of three experiments. In these experiments, the ratio of the IC, for IL-lra to the IC, for ILla ranged from 250 to 533. were incubated overnight at 4 "C with -100 pM '"I-IL-1s as well as increasing concentrations of IL-1B (0) or IL-Ira (0). Nonspecific binding represented 10% of total binding.  4-6), or EL4 cells (lanes 7-9) were incubated overnight a t 4 "C with -1 nM "'I-IL-ls alone (lanes I, 4, and 7) or in the presence of 1 p M IL-IP (lanes 2, 5, and 8) or 10 p~ IL-Ira (lanes 3, 6, and 9). Following cross-linking with BSa, the IL-lP-IL-IR complex was immunoprecipitated with anti-IL-lD and then analyzed by SDS-PAGE.

and EL4 cells. PMN (lanes I-3), Raji cells (lanes
IL-lP. IL-1RtII complex was then immunoprecipated using rabbit anti-human IL-1P antibody before being analyzed by SDS-PAGE. As shown in Fig. 5, cross-linking of radiolabeled IL-lB to PMN and Raji cells shows a band at 85 kDa. Conversely, cross-linking of radiolabeled IL-lP to EL4 cells bearing IL-1RtI resulted in a band at 97 kDa. In the presence of an excess of either unlabeled IL-lP or IL-lra, neither the 85-kDa IL-1. IL-1RtII complex nor the 97-kDa IL-1-IL-1RtI is detected. Similar results were obtained when PMN, Raji cells, and EL4 cells were incubated with 1 nM '2sI-IL-1a in the presence or absence of unlabeled IL-la or IL-lra (data not shown).

DISCUSSION
Natural human urine-derived IL-Ira (formerly called the IL-1 inhibitor) has been shown to block the binding of human IL-1 to murine EL4-6.1 thymoma cells (17). After cloning this molecule, Hannum et al. (2) and Carter et al. (1) found that the recombinant protein could competitively inhibit the binding of human lzSI-IL-la to both EL4 cells and human NK-like YT cells possessing the IL-1RtI. In the murine cells, IL-la was a more potent inhibitor than IL-lra; in the human cells IL-la and IL-Ira were equipotent as competitive inhibitors. However, in neither report was IL-lra at concentrations up to 50 nM able to block the binding of 3sS-labeled IL-la (2) or '2sI-labeled IL-la (1) to 70Z/3 cells, a murine pre-B-cell line which expresses the IL-1RtII.
Data presented in this paper show that IL-lra competitively inhibits the binding of '251-IL-1/3 to cells with the IL-1RtII. IL-1P is 1 order of magnitude more potent than IL-lra in inhibiting the binding of '251-IL-1P to PMN and Raji human B-lymphoma cells. In further support of the theory that IL-1P and IL-lra bind to the same receptor, IL-1P is 1 order of magnitude more potent than IL-Ira in blocking the binding of '2sI-IL-lra to PMN. Cross-linking studies confirm that IL-Ira is able to prevent '2sI-IL-1/3 from binding to a 68-kDa receptor.
Species specificity probably is the reason previous studies did not demonstrate inhibition of IL-1 binding by human IL-Ira to the mouse IL-1RtII. We believe that the affinity of the mouse IL-1RtII for mouse IL-Ira would be greater than its affinity for human IL-lra. Ikejima' has shown that mouse EL4-6.1 cells have a higher affinity for mouse '2sI-IL-1P than human '2sI-IL-l@.