Abstract
IL-18 is a pleiotropic proinflammatory cytokine that is involved in induction of inflammatory mediators, regulation of the cytotoxic activity of NK cells and T cells, and differentiation and activation of both Th1 and Th2 cells. IL-18 signals through its specific cell surface receptor IL-18R, which comprises two subunits: IL-18Rα and IL-18Rβ. IL-18Rα alone has a weak affinity for IL-18 binding, while the IL-18Rα/β complex has a high affinity. By using several anti-IL-18 mAbs and IL-18 binding protein, we have examined whether these site-specific inhibitors could block the binding of IL-18 to IL-18Rα and to the IL-18Rα/β complex. Here we show that IL-18 binding to IL-18Rα was inhibited by a neutralizing mAb, 125-2H, while binding of IL-18 to the α/β receptor complex was not. This suggests that IL-18Rβ-induced conformational changes may occur in IL-18Rα upon dimerization, leading to changes in the presentation of IL-18 binding sites. Epitope mapping of 125-2H using human-mouse IL-18 chimeras identified a region in IL-18 that was required for 125-2H recognition. This region, as examined by IL-18R binding and functional analysis, appeared to be critical for triggering signal transduction through the heterodimeric receptor.
Interleukin-18, initially defined as an IFN-γ-inducing factor, is a proinflammatory cytokine that belongs to the IL-1 cytokine family (1). Like IL-1β, IL-18 is expressed as an inactive pro-peptide (proIL-18) and is processed by caspase-1 to form an active mature protein (2, 3). It has been shown to have a broad range of effector functions beyond lymphocyte activation that implicate IL-18 as an important regulator of both innate and acquired immunity (4, 5). IL-18 expression is elevated at sites of chronic inflammation in human autoimmune diseases (6, 7, 8), and blockage of its activity showed efficacy in animal models of rheumatoid arthritis (7), multiple sclerosis (9), colitis (10), and liver disease (11). Hence, IL-18 is a potential drug target for the treatment of inflammatory indications.
The binding of IL-18 to its target cells is mediated by specific cell surface receptors, which are similar to the IL-1R system. The receptor of IL-18 is composed of an α-chain (IL-18Rα) (12, 13) and a β-chain (IL-18Rβ or AcPL) (14). The α-chain of the IL-18R alone can bind IL-18 with an affinity of 18.5 nM (12) and does not signal (14). Like the IL-1R accessory protein of the IL-1R complex, IL-18Rβ does not bind IL-18 alone, but forms a functional high affinity (0.4 nM) receptor complex with IL-18Rα that is able to signal in response to IL-18 (15). It has been proposed that IL-18Rβ does not directly interact with IL-18, and it is IL-18Rα that is solely responsible for IL-18 binding. Given the difference in affinities between IL-18Rα and IL-18Rα/β complex for IL-18 binding, it is likely that IL-18Rα and IL-18Rα/β complex may present different contact sites for IL-18. These differences may involve conformational changes, leading to different orientations as well as different numbers of contacting sites. Our current investigation, through the use of site-specific antagonists, shows that a mAb is able to efficiently block IL-18 binding to IL-18Rα, but not to the IL-18Rα/β complex. Interestingly, this Ab is still able to effectively neutralize IL-18-mediated signaling transduced through the IL-18Rα/β complex. We have mapped the epitope of this Ab and showed that the C-terminal 17-aa sequence of human IL-18 is critical for signal transduction through the heterodimeric receptor, but is not required for binding to the same receptor. Our data could provide valuable insights regarding the complex mechanism of IL-18 binding to its receptor heterodimer.
Materials and Methods
Determination of IL-18R expression in L428 and KG-1 cells
Expression of IL-18R subunits was studied using human Hodgkin’s lymphoma cell line L428 (obtained from Hayashibara Biochemical Laboratories, Okayama, Japan), human myelomonocytic cell line KG-1 (purchased from American Type Culture Collection, Manassas, VA), and KG-1 cells pretreated with 10 ng/ml recombinant human TNF-α (rhTNF-α 2
IL-18R binding assay
7 cells/ml in RPMI. The viability of the cell cultures, as determined by vital dye staining with 0.4% trypan blue stain, was 95% in the binding experiments. To examine the potency of mAbs to inhibit IL-18R binding, 100 μl of the cell suspension (5 × 106 cells/well) was incubated with a constant amount of 125I-labeled rhIL-18 (2 nM) with or without serial dilutions of either human IgG1 or anti-IL-18 Abs/IL-18BP in 96-well, U-bottom, microtiter plates for ∼4 h on ice in a final total volume of 0.2 ml. The level of nonspecific binding in each experiment was determined by adding a 2800-fold excess of unlabeled rhIL-18 to control wells. Thereafter, 75 μl of the mixture was transferred to 1.0-ml test tubes containing a 300-μl mixture of dibutylphthalate (Sigma-Aldrich, St. Louis, MO) and dinonylphthalate (ICN Pharmaceuticals, Costa Mesa, CA) at a 2:1 volume ratio. Free 125I-labeled rhIL-18 was removed by centrifugation for 5 min. Each test tube end containing the cell pellet was cut with microtube scissors. The pellet contained cell-bound 125I-labeled rhIL-18, and the aqueous phase above the phthalate oil mixture contained excess free 125I-labeled rhIL-18. Cell pellets were collected in counting tubes, and cell-bound radioactivity was determined using a Packard Cobra gamma counter. All results were determined in duplicate and expressed as the average. Each experiment was repeated three times. The data shown are from one representative experiment. Data are presented as the percent inhibition of total 125I-labeled rhIL-18 binding to the L428 or rhTNF-α-stimulated KG-1 cells against the concentrations of rhIL-18 antagonists. Fifty percent inhibitory concentration (IC50) values were then calculated with a nonlinear, four-parameter, sigmoidal curve-fitting method using Origin software version 6.0 (Microcal Software, Northampton, MA).
Analysis of IL-18-induced IFN-γ production by human KG-1 cells
IL-18 bioassay using KG-1 cells was performed as described previously (16). Briefly, rhIL-18 preparations (in a final concentration of 2 ng/ml) were added to KG-1 cells (3 × 106/ml) in RPMI medium containing 10 ng/ml hTNF-α and incubated at 37°C for 16–20 h. For the Ab neutralization assay, IL-18 samples (in a final concentration of 2 ng/ml) were preincubated with mAb (in final concentrations between 0 and 10 μg/ml) or IL-18BP (in final concentrations between 0 and 1 μg/ml) at 37°C for 1 h and then added to KG-1 cells (3 × 106
Surface plasmon resonance
Real-time binding interactions between captured Ab (mouse anti-rhIL-18 Ab captured on a biosensor matrix via goat anti-mouse IgG, and IL-18 8P/Fc was captured by goat anti-human Fc) and rhIL-18 were measured by surface plasmon resonance using the BIAcore system (BIAcore, Piscataway, NJ) according to the manufacturer’s instructions and standard procedures. Briefly, IL-18 was diluted in HBS running buffer (BIAcore), and 50-μl aliquots were injected through the immobilized protein matrixes at a flow rate of 5 μl/min. The concentrations of rhIL-18 used were 62.5, 125, 187.5, 250, 375, 500, 750, 1000, 1500, and 2000 nM. To determine the dissociation constant (off-rate), association constant (on-rate), and KD, BIAcore kinetic evaluation software (version 3.1) was used.
Generation of IL-18 mutant proteins
A series of human-mouse proIL-18 chimeric constructs were generated by overlapping PCR using human and mouse proIL-18 cDNA as templates, and the final PCR products were subcloned into pcDNA3.1 TOPO vector (Invitrogen, San Diego, CA). The plasmids, each containing a proIL-18 wild-type (wt) or mutant construct, were used as templates for in vitro transcription and translation (TNT) using the TNT Quick Coupled System (Promega) according to the manufacturer’s instructions. Briefly 1 μg of DNA and 1 μl of methionine were added to 40 μl of reaction mixture and incubated at 30°C for 90 min. The resulting protein products were processed into the mature form of IL-18 by caspase-1 digestion at 25°C for 1 h. The wt human and mouse IL-18 processed using the above procedure was quantified by ELISA (17). To quantify the pro and mature forms of mutant IL-18 proteins, [35S]methionine was used in parallel TNT reactions. Before and after caspase-1 processing, wt and mutant samples were applied on SDS-PAGE and quantified densitometrically using a Storm 860 image analysis system (ApBiotech). The biological activity of each mutant protein was analyzed by stimulating TNF-α-primed KG-1 cells with each mutant protein at concentrations in the range of 0.1–50 ng/ml, and resulting IFN-γ production was measured by ELISA. The activities of wt and mutant IL-18 proteins at 2 ng/ml, which was within the linear range of the dose-response curve, are shown.
125-2H binding to mutant IL-18 proteins
wt or mutant IL-18 TNT preparations (100 ng/ml) were added to the wells and incubated overnight at 4°C, followed by detection with 125-2H (2 μg/ml) and HRP-anti-mouse IgG Fc (human IgG absorbed; Sigma-Aldrich) as the secondary Ab. As a control, another anti-hIL-18 mAb, mAb318, which was reactive with all the mutants, was also used as a detecting Ab to show proper capturing of all mutant proteins by IL-18BP. Following substrate incubation, samples were read using a microplate reader (Molecular Devices, Sunnyvale, CA).
Results
Expression of IL-18Rα and β subunits and binding of IL-18
IL-18R is expressed on a variety of cell types, including T and B cells and myeloid, monocytoid, erythroid, and megakaryocytic cell lines. Among the many cell lines tested, L428 exhibited the strongest level of total IL-18 binding, but failed to secret IFN-γ in response to IL-18 (12, 18), and KG-1 cells produced IFN-γ in response to IL-18 stimulation (18). However the IL-18R expression patterns (i.e., whether they express only IL-18Rα or both IL-18Rα and -β, and the receptor numbers and affinities) on these cells have not been assessed in detail. We have examined these two cell lines with regard to the level of IL-18Rα/β expression using flow cytometry and receptor binding assay. Previous data showed that TNF-α could significantly increase the response of KG-1 cells to IL-18 in IFN-γ production, suggesting that TNF-α might up-regulate the expression of IL-18Rα and/or IL-18Rβ (18). Therefore, we also analyzed TNF-α-primed KG-1 cells for receptor expression and IL-18 binding. The flow cytometric data showed that both L428 cells and KG-1 cells express IL-18Rα (Fig. 1⇓). Resting KG-1 cells expressed much less IL-18Rα than L428 cells; however, upon TNF-α stimulation, IL-18Rα expression in KG-1 cells was significantly up-regulated (Fig. 1⇓). Our results also showed that both L428 cells and KG-1 cells expressed low levels of IL-18Rβ, and the level of IL-18Rβ expression was significantly increased in TNF-α-primed KG-1 cells (Fig. 1⇓). The expression of both IL-18Rα and -β on L428 cells was not altered when cells were treated with TNF-α, since these cells did not express either p55 or p75 of TNF receptors as measured by flow cytometry (data not shown). Based on the above observations, we concluded that L428 cells primarily expressed low affinity IL-18Rα and did not respond to IL-18, while TNF-α-primed KG-1 cells expressed high affinity IL-18Rα/β heterodimer and responded to IL-18 as measured by IFN-γ production. We therefore used these two cell lines to analyze IL-18 binding to IL-18Rα and the IL-18Rα/β complex.
Expression of IL-18R on L428 and KG-1 cells. IL-18Rα and IL-18Rβ expressions were analyzed on L428, KG-1, and TNF-α-primed KG-1 (pretreated with 10 ng/ml TNF-α for 24 h) cells by flow cytometry. The y-axis shows the cell counts, and the x-axis is the signal intensity of PE-labeled anti-IL-18Rα and FITC-labeled rabbit anti-mouse IgG detecting mouse anti-hIL-18Rβ.
Neutralizing mAb 125-2H blocks IL-18 binding to IL-18Rα effectively, but not to the IL-18Rα/β complex
Using three different neutralizing mAbs and IL-18BP in a receptor binding assay, we examined the interactions of IL-18 with IL-18Rα and with the IL-18Rα/β complex. Such information will allow identification of specific regions of IL-18 that might be involved in the process during which a low affinity binding to IL-18Rα is transformed into a high affinity binding to IL-18Rα/β complex. Our data showed that all three neutralizing Abs and IL-18BP were able to specifically block IL-18 binding to L428 cells (Fig. 2⇓A). IL-18BP was the most potent inhibitor, followed by mAbs 125-2H, Mab318, and B-K22 (Fig. 2⇓A and Table I⇓). In TNF-α-primed KG-1 cells, the potencies of IL-18BP, Mab318, and B-K22 in blocking IL-18 binding were similar to those seen in L428 cells (Fig. 2⇓B and Table I⇓). However, the inhibitory potency of 125-2H in TNF-α-primed KG-1 cells was decreased dramatically (IC50 = >1000 nM) compared with that in L428 cells (IC50 = 1.78 nM; Table I⇓). Interestingly, 125-2H was still very potent in neutralizing IL-18 biological activity in TNF-α-primed KG-1 cells (Fig. 3⇓ and Table I⇓). The neutralizing potencies of the Abs seen in the KG-1 bioassay correlated reasonably well with their affinities for IL-18 as determined by BIAcore analysis (Table I⇓). 125-2H displayed highest affinity among all Abs tested, roughly 10- and 100-fold higher than Mab318 and B-K22, respectively. IL-18BP also exhibited high affinity to IL-18. Our data indicate that 125-2H could effectively block IL-18 binding to the α subunit of the receptor (on L428 cells), but not to the IL-18Rα/β complex (on TNF-α-primed KG-1 cells). Nevertheless, 125-2H could inhibit IL-18 biological activity in the same TNF-α-primed KG-1 cells with high potency. Collectively, these data suggest that, at least at the region of 125-2H epitope, IL-18 interacts with IL-18Rα and IL-18Rα/β complex quite differently. The 125-2H epitope region of IL-18 might be critically involved in the interaction with IL-18Rα/β complex, leading to signal transduction, but is not required for binding of IL-18 to the receptor complex. To define this particular region, we defined the binding epitope of 125-2H.
Inhibition of IL-18R binding by anti-IL-18 mAbs and IL-18BP. Inhibition of binding of 125I-labeled rhIL-18 to L428 (A) or TNF-α-primed KG-1 (B) cells by anti-IL-18 mAbs and IL-18BP was determined as described in Materials and Methods. Data are presented as the percent inhibition vs increasing concentrations of IL-18 Abs or IL-18BP.
Effects of anti-IL-18 mAbs and IL-18BP on IL-18-mediated IFN-γ production by primed KG-1 cells. TNF-α-primed KG-1 cells were stimulated at 37°C for 16–20 h by rhIL-18 (2 ng/ml) that had been preincubated for 1 h at 37°C with varying concentrations of mAbs or IL-18BP, and IFN-γ secretion in the medium was quantified by ELISA. Results are presented as the percentage of total IFN-γ production in the absence of inhibitors.
Characterizations of IL-18R binding using mAb and IL-18BPa
Generating IL-18 chimera mutants for mapping the 125-2H epitope
Previous attempts to map the 125-2H epitope using overlapping peptides were unsuccessful (C. Wu, unpublished results), suggesting that 125-2H may be against a conformational and nonlinear epitope. The intact three-dimensional structure of IL-18 appears to be required for 125-2H recognition. We have developed a functional approach to delineate the binding epitope of 125-2H using human-mouse IL-18 chimera mutants as shown in Fig. 4⇓. Since 125-2H does not recognize mouse IL-18 (data not show), and human and mouse IL-18 share sequence and structural homologies, we replaced various regions of human IL-18 with the corresponding mouse IL-18 sequence so that the proportion of human sequence was gradually increased and the mouse sequence decreased, allowing a region specific for 125-2H binding in human IL-18 to be identified. We first determined whether the chimera IL-18 mutants HaMb1–4 were biologically active. This was done by stimulating KG-1 cells with wt or mutant IL-18 proteins (0–50 ng/ml) in the presence of TNF-α, and the secreted IFN-γ was quantified by ELISA. A dose-response curve was generated to assess the biological potency of each mutant. Only results of wt and mutant IL-18 at a 2 ng/ml concentration are shown (Fig. 5⇓). Mouse IL-18 displayed only minimal activity on human KG-1 cells in inducing IFN-γ production (170 vs 82 pg/ml background signal), and human IL-18 (1120 pg/ml IFN-γ, as shown in Fig. 5⇓) is 5- to 10-fold more potent than mouse IL-18 in this assay. Our data showed that all the chimera mutants were biologically functional at inducing IFN-γ secretion by KG-1 cells, although their potencies were different (Fig. 5⇓). Strikingly HaMb1 exhibited a 7-fold higher potency than wt rhIL-18, whereas HaMb2, -3, and -4 showed activities similar to that of rhIL-18 (Fig. 5⇓).
Schematic diagram of IL-18 chimera constructs and their expression. IL-18 human-mouse chimera constructs were generated by overlapping PCR and were produced by in vitro TNT, followed by caspase-1 processing; results were quantified by ELISA and SDS-PAGE. □, human proIL-18 sequence; ▪, mouse IL-18 sequence, with amino acid residue numbers indicated within each box. Mutants HaMb1–4 (A, upper panel) and their quantification on radioactive SDS-PAGE (A, lower panel) are shown. Both processed (mature IL-18) and unprocessed (pro IL-18) proteins are indicated by arrows. B, Upper panel, Constructs of mutant HaMb1hc, HaMb3hc, and HaMb6. They were processed and quantified similarly, as shown in the lower panel.
Biological activity of IL-18 mutants on human KG-1 cells. Human IFN-γ production by TNF-α-primed KG-1 cells. Cells were pretreated with 10 ng/ml TNF-α, followed by incubation with varying concentrations of wt and mutant IL-18 proteins for 16–20 h. Supernatants were collected for IFN-γ quantification by standard ELISA. The figure shows the activities of mutants at a 2 ng/ml concentration.
125-2H binding requires C-terminal 17 aa of IL-18 in correct conformation
Since our first four mutants were biologically functional, we were able to analyze their interactions with 125-2H Ab in a functional assay. Increasing concentrations of 125-2H were preincubated with wt or mutant IL-18 preparations and added to KG-1 cells to induce IFN-γ production. In this neutralization assay none of the four mutants (constructs shown in Fig. 4⇑A) was neutralized by 125-2H, while the activity of wt rhIL-18 was inhibited by 125-2H in a dose-dependent manner (Fig. 6⇓A). The data suggested that the last 17 aa residues of human IL-18 were required for 125-2H recognition, since this region represented the only difference between wt rhIL-18 and mutant HaMb4. Based on this hypothesis, additional mutants were generated, as shown in Fig. 4⇑B, where the last 17 residues of HaMb1 and HaMb3 were replaced by the human counterpart to generate HaMb1hc and HaMb3hc, respectively, or the last 17 residues of human IL-18 were added to HaMb1 at its C terminus to generate HaMb6. This was to determine whether this 17-aa sequence, when placed within the context of the IL-18 molecule or outside the molecule as an attachment, would render the chimera molecules recognizable to 125-2H. All three mutants were biologically active (Fig. 5⇑). Strikingly, mutant HaMb6 displayed 10-fold higher potency than wt hIL-18 in the KG-1 assay (Fig. 5⇑). The results of the neutralization assay also showed that while 125-2H could effectively inhibit the activity of both HaMb1hc and HaMb3hc in a dose-dependent manner, it did not block the function of HaMb6 (Fig. 6⇓B). An ELISA-based binding assay also showed that HAMb6 did not bind 125-2H, but both HaMb1hc and HaMb3hc did (Fig. 7⇓). These results collectively demonstrate that the C-terminal 17-aa sequence of IL-18 is indispensable for 125-2H binding, and that this sequence requires a correct conformation within the tertiary structure of IL-18 for 125-2H recognition.
The C-terminal 17-aa sequence of rhIL-18 is required for 125-2H recognition. Inhibition of the activities of wt and mutant IL-18 proteins by 125-2H was analyzed using the KG-1 assay. IFN-γ production by TNF-α-primed KG-1 cells was measured by ELISA following stimulation by wt or IL-18 mutants, HaMb1 to HaMb4 (A) and HaMb1hc, HaMb3hc, and HaMb6 (B) at 2 ng/ml in the presence of various concentrations of 125-2H. Results are presented as the percentage of total IFN-γ production in the absence of 125-2H.
Binding of IL-18 mutants to 125-2H Ab by ELISA. Binding of 125-2H to wt and mutant IL-18 proteins was analyzed by ELISA using IL-18BP-Fc as the capture reagent. Upper panel, 125-2H was used as the detecting Ab; lower panel, Mab318, which binds to all mutants, was used as the detecting Ab as a control. Equal amounts (100 ng/well) of wt and mutant IL-18 proteins were used for both ELISAs. An HRP-labeled rabbit anti-mouse Ig (human absorbed) was used as the secondary Ab.
Discussion
The IL-1 cytokine family members, including IL-1α, IL-1β, IL-1R antagonist, IL-18, and several IL-1 homologues, share similar structural folds. Likewise, their receptor systems are also similar, comprising of a ligand-binding subunit (α-chain) and a signaling subunit (β-chain). Since β-chain alone does not measurably bind ligand, it is widely accepted that the α-chain is solely responsible for ligand binding. Extensive studies have been performed to understand the molecular interactions between IL-1 and IL-1Rα, and crystal structures of the complexes of IL-1Rα/IL-1β and IL-1Rα/IL-1R antagonist have been reported (19, 20). However, the role of the receptor β-chain in the process of ligand-receptor complex interactions of IL-1 family cytokines has not been extensively evaluated.
IL-18Rα alone can bind IL-18, but does not signal, and coexpression of IL-18Rβ results in functional reconstitution of the IL-18R complex (21), exhibiting a significantly increased binding affinity for IL-18 (12, 14, 22). In TCR-triggered T cells, IL-12 can generate functional IL-18R complex through induction of IL-18Rβ expression as well as up-regulation of the IL-18Rα in a STAT4-dependent manner (23). Overexpression of IL-18Rβ is observed in lymph node cells of autoimmune MRLlpr/lpr mice, whereas IL-18Rα expression is normal. As a result, these cells are hyper-reactive to IL-18 stimulation in terms of both IFN-γ production and cell proliferation (24). These reports collectively show that IL-18Rβ is a crucial regulatory element in the cellular response to IL-18 and is critical for thorough understanding of the interactions between IL-18 and the functional IL-18Rα/β heterodimer.
In the present study we have demonstrated that the binding of IL-18 to IL-18Rα and that to the IL-18Rα/β complex appear to be different through the use of site-specific IL-18 inhibitors, including three mAbs and IL-18BP. All four inhibitors could block the binding of IL-18 to IL-18Rα on L428 cells, but only three (two Abs and IL-18BP) blocked the binding of IL-18 to IL-18Rα/β complex on TNF-α-primed KG-1 cells. The high affinity Ab, clone 125-2H, was not able to efficiently block IL-18 binding to IL-18Rα/β complex on TNF-α-primed KG-1 cells. This could not be due to IL-18 binding to an unknown receptor on KG-1 cells, since an anti-IL-18Rα Ab could completely block IL-18 binding to KG-1 cells in both the presence and the absence of 125-2H (data not shown). Interestingly, on the same TNF-α-primed KG-1 cells, 125-2H was a very potent neutralization agent inhibiting IL-18-mediated IFN-γ production. These observations have prompted us to propose the following. 1) The C-terminal 17-aa sequence of IL-18 might be critically involved in interactions with monomeric IL-18Rα, and therefore 125-2H can effectively block IL-18 binding to IL-18Rα. 2) Upon receptor dimerization and formation of the IL-18Rα/β complex, conformational changes take place in IL-18Rα that may lead to changes in the interactions with IL-18. Specifically the C-terminal region of IL-18 is probably now displaced from the IL-18Rα binding site and therefore renders 125-2H ineffective in blocking the binding of IL-18 to IL-18Rα in the heterodimeric format. 3) It is intriguing that 125-2H can inhibit signaling, but not binding, of IL-18 to IL-18Rα/β. We speculate that 125-2H may block a critical region in IL-18 that is necessary for signal transduction and could be a region that interacts with the receptor β-chain. 4) Our data also suggested that dimerization of receptors α and β occurs before (or independent of) ligand binding, since 125-2H could block IL-18 binding to IL-18Rα, but not to the IL-18Rα/β complex.
Our results from 125-2H epitope mapping indicate that the C-terminal 17-aa sequence of IL-18 is required for 125-2H recognition. We have mapped this epitope through a functional approach by analyzing chimera human-mouse IL-18 mutant proteins in the KG-1 assay. All chimera mutant IL-18 proteins were biologically active, some of which displayed significantly increased potencies compared with wt IL-18. This may be due to the increased binding affinity for the receptor. Among all mutants analyzed, only the two mutants (HaMb1hc and HaMb3hc) that possessed the intact C-terminal 17-aa sequence of human IL-18 were neutralized by 125-2H. The IC50 values for these two mutants were higher than that for wt IL-18, probably due to increased affinity to IL-18R or decreased affinity for 125-2H. Nevertheless, they were both neutralized by 125-2H in a dose-dependent manner. An ELISA-based assay also showed that 125-2H bound to these mutants as well. These data elucidated that the C-terminal 17-aa residues of rhIL-18 were of central importance in 125-2H recognition. In addition, this 17-aa sequence required a correct conformation in the context of IL-18 tertiary structure for 125-2H recognition, demonstrating that the binding epitope of 125-2H was probably a conformational epitope. In the published IL-1β/receptor structure (19), the C-terminal sequence is positioned in the interface between IL-1 and IL-1Rα, in agreement with the idea that this part of the molecule was involved in direct contact with receptor α-chain. In Fig. 8⇓, a possible mechanism by which 125-2H and IL-18 interact in the two cellular systems is shown. Because of the IL-18Rβ-induced changes in IL-18Rα, the epitope of 125-2H is no longer directly involved in receptor binding; instead, it may be critical for IL-18Rβ-mediated signaling.
Diagram of IL-18 binding model on L428 and KG-1 cells. The left half of the diagram represents 125-2H inhibition of IL-18 binding to L428 cells expressing only IL-18Rα. The white circle within the IL-18 molecule represents binding epitope of 125-2H. Since 125-2H binding epitope was directly involved in receptor binding, the interaction of IL-18 with the receptor was abolished in the presence of 125-2H. In primed KG-1 cells, shown on the right half of the picture, the presence of IL-18Rβ may induce a conformational change in IL-18Rα, resulting in displacement of 125-2H epitope out of the binding pocket. Therefore, the interaction of IL-18 with the receptor complex was not inhibited by 125-2H. However 125-2H epitope might be involved in interacting with IL-18Rβ, which could be critical for signal transduction. As a result, 125-2H can inhibit IL-18-induced IFN-γ production by KG-1 cells.
The C-terminal region of IL-1β has been shown to be involved in receptor interaction through studies using mAbs and peptides (25, 26, 27). The crystal structure of IL-1β/IL-1R has indicated that the C-terminal region of IL-1β fits deeply into the binding pocket between Ig-like domain 2 and domain 3 of the α-chain receptor (19). Due to the unique position of the C terminus, residues at this region contribute to both receptor binding sites A and B of IL-1β (19). Given the similar overall structural fold shared among IL-1 family members and based on our results from IL-18, we predict that the involvement of the C-terminal region in receptor interactions is also common throughout the family. We speculate that the receptor signaling subunit-induced binding site changes in the ligand binding subunit of the receptor may also be shared among IL-1 family receptors.
Acknowledgments
We thank Dr. Yong-In Kim for supplying the caspase-1, and Drs. David Presky, Robert Talanian, Subhashis Banerjee, Roseanne Waterhouse, Jochen Salfeld, Ken Zwicker, Zehra Kaymakcalan, and Andrea Pellacani for critical review of the manuscript.
Footnotes
↵1 Address correspondence and reprint requests to Dr. Chengbin Wu, Abbott Bioresearch Center, 100 Research Drive, Worcester, MA 01605. E-mail address: chengbin.wu{at}abbott.com
↵2 Abbreviations used in this paper: rh, recombinant human; IC50, 50% inhibitory concentration; IL-18BP, IL-18 binding protein; TNT, transcription and translation; wt, wild type.
- Received March 20, 2003.
- Accepted March 31, 2003.
- Copyright © 2003 by The American Association of Immunologists
References
In this issue
