J Biol Chem, Vol. 274, Issue 29, 20103-20109, July 16, 1999

Recruitment of a Heparan Sulfate Subunit to the Interleukin-1 Receptor Complex
REGULATION BY FIBRONECTIN ATTACHMENT^*

Soraya Vallés, Christina Tsoi^§, Wen-Yan Huang, David Wyllie, Franco Carlotti, Janet A. Askari^¶, Martin J. Humphries^¶, Steven K. Dower, and Eva E. Qwarnström^§

From the  Functional Genomics Group, Division of Molecular and Genetic Medicine, Royal Hallamshire Hospital, University of Sheffield, Glossop Rd., Sheffield S10 2JF, United Kingdom, the ^§ Department of Pathology, University of Washington, Seattle, Washington, and the ^¶ Wellcome Trust Center for Cell-Matrix Research, School of Biological Sciences, University of Manchester, Manchester M13 9PT, United Kingdom

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

In this study, we identified an adhesion-regulated subunit of the interleukin-1 (IL-1) receptor complex. Transfection of fibroblasts with an IL-1 receptor-EGFP construct showed that the fusion protein was located at focal adhesions in cells attaching to fibronectin.

Fibronectin attachment caused enhancement in endogenous IL-1 type I receptor levels from on average 2500 to 4300 receptors/cell. In addition, matrix attachment resulted in a decrease in binding affinity (K_a) from 1.0 × 10⁹ (M¹) to 5.6 × 10⁸ (M¹), due to a 2-fold reduction in association rate constant.

The adhesion-mediated effects were reversed by soluble heparin. Cross-linking experiments showed that in cells attached to fibronectin, 50-70% of the radiolabeled IL-1 was associated with a heparinase sensitive, high molecular mass component of about 300 kDa, with a core protein of 80-90 kDa. Formation of the complex was dependent on cell interaction with the heparin binding region in fibronectin and required IL-1/type I IL-1 receptor binding.

This report demonstrates the recruitment of a heparan sulfate to the IL-1 receptor complex, following attachment to fibronectin, which correlates with alterations in receptor function. The data suggest that the heparan sulfate constitutes an attachment regulated component of the IL-1 receptor complex with the role of mediating matrix regulation of IL-1 responses.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Regulation of a number of cytokine and growth factor receptors is mediated in part through accessory proteins, recruited before or following ligand binding. In several instances, such as modulation of fibroblast growth factor-receptor interaction and IL-6 signaling, regulation involves interactions with glycoproteins and proteoglycans (1, 2).

IL-1 responses are regulated through a receptor complex to which an increasing number of components have been found to belong. The receptor uses a well conserved mechanism for regulating immune and inflammatory responses (3-7) but which is still not completely understood. Initiation of the signal occurs through a heterodimeric receptor complex, composed of the interleukin type I receptor (IL-1RI)¹ (8) and the accessory protein (AcP) (9), both 80-90 kDa glycoproteins. The type I receptor complex is present on T cells and connective tissue derived cells (8, 10) characteristically at low levels (11). The receptor binds both forms of IL-1 (12, 13), with active forms of 17 kDa, and is responsible for IL-1 induced activation of both NF-B and stress kinase pathways (14-17). In addition, a 67-kDa, nonsignaling receptor (type II) is expressed on B cells, monocytes, and T cells (18-22). The receptor proximal stages of signal activation involve the adapter proteins TRAF 6 (23) and MyD88 (24-26) and two related kinases, IRAK-1 and IRAK-2 (7, 24).

In adherent cells, such as fibroblasts, IL-1 receptors are located at focal adhesions (27, 28), and IL-1 binding to the type I receptor has rapid effects on cell structure (29). IL-1 signal transduction is regulated by cell attachment and spreading (30), and fibronectin attachment is permissive in IL-1 responses in adherent cells (31).

In this report we demonstrate the presence of a novel, attachment-regulated component of the IL-1 receptor complex at focal adhesions that constitutes a heparan sulfate proteoglycan. Recruitment of this component affects receptor function, both in terms of the level of type I IL-1 receptors and their affinity for ligand, and correlates with IL-1-mediated, attachment-regulated signaling and biological responses.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Transfection-- Cells grown in 10-cm dishes (3 × 10⁶cells/dish) were transfected with a construct derived from pEGFP-N2 (CLONTECH, Palo Alto, CA), in which the full-length IL-1RI was introduced into the vector mcs upstream of the EGFP sequence. The receptor portion, a chimera (HinMext) containing the mouse extracellular domain linked to the human intracellular domain (32), was excised from the vector pDC406 using XhoI and BsaAI. The vector p-EGFP-N2 was digested with XhoI and SmaI, and the IL-1 receptor, lacking the C-terminal 4 residues (VPLG), was inserted in frame in front of a sequence encoding the linker (GIHRPVAT). Transfection was carried out by calcium phosphate precipitation (4 µg of plasmid/10-cm dish for 4 h at 37 °C, followed by 25% Glycerol shock for 1 min). Twenty-four hours after transfection, cells were detached using EDTA (5 mM), replated on tissue culture treated multiwell chambered coverglasses (Labtek, Nunc) coated with fibronectin (10 µg/ml) as below, and allowed to attach for 4 h.

Immunostaining-- Transfected cells, replated on fibronectin for 4 h, as below, were fixed for 10 min in PBS + 4% paraformaldehyde, washed three times with PBS, and blocked/permeabilized by incubation in PBS + 5% goat serum + 0.1% Triton X-100. Primary mouse anti-vinculin antibody (V9131, Sigma) (overnight at 4 °C) and secondary biotinylated goat anti-mouse IgG antisera (Santa Cruz Biotechnology Inc., Santa Cruz, CA) (30 min at room temperature) were applied sequentially at 2 µg/ml in blocking/permeabilization buffer followed by Texas Red-conjugated streptavidin (Molecular Probes) (20 min at room temperature) at 0.2 µg/ml in PBS. After extensive washing, images of vinculin staining (red fluorescence) and IL-1RI-EGFP (green fluorescence) were obtained by confocal microscopy.

Confocal Microscopy-- Data were acquired with a Multiprobe CLSM 2010 laser scanning confocal microscope (Molecular Dynamics) interfaced to a Silicon Graphics workstation, using the 488 nM line (excitation for green fluoresence) and the 565 nM line (excitation for red fluorescence) of a krypton/argon mixed gas laser and a 50-µm pinhole aperture, generating a 0.5-µm-thick optical section from a planApo × 60 oil lens (NA 1.40). Data were collected using a 530 nM band pass filter (green) and a 590 nM long pass filter (red) and at a photo multiplier tube voltage in the range of 550-800 V and 10% laser attenuation. After acquisition, data files were exported to a Power Macintosh, and converted to 8 bit TIFF images using NIH Image (Wayne Rasband at the National Institutes of Health and available from the Internet) for printing and analysis.

Tissue Culture-- Human dermal or gingival fibroblasts (six strains, transfer 9-14) were used. Both types of fibroblasts were tested in all assays, and they consistently showed the same effects. Cells were plated in 6-well (2 × 10⁵ cells/well) or 24-well (5 × 10⁴ cells/well) tissue culture plates, on bare tissue culture plastic, or in wells coated with fibronectin (10 µg/ml) (Life Technologies, Inc.) or on poly-L-lysine (0.1, 0.5, and 1.0 mg/ml) (Sigma) (33, 16) following detachment using EDTA (5 mM), as described (31), and allowed to attach for 3-4 h in 10% fetal calf serum. Using poly-L-lysine resulted in the same responses as those seen in cells on bare tissue culture plastic, shown in all experiments together with fibronectin. In some experiments, cells were incubated prior to plating (5 min) and plated in the presence of heparin (10⁹-10⁵ M) or with the same concentration of chondroitin sulfate, as control, or in the presence of an RGD containing peptide (GRGDSP, 10⁸-10⁴ M, Life Technologies, Inc.) or a nonspecific, RGE-containing peptide (GRGESP, 10⁸-10⁴ M, Life Technologies, Inc.) used as a control. Furthermore, in some experiments, cells were plated on a wild type or a mutant version of the H/95 recombinant fibronectin fragment containing the HepII/IIICS region (34). The mutant protein was altered in the main heparin binding region in HepII (PPRRARVT to PPSSASVT) and in an additional site in IIICS (HGFRRTTP to PPSSASVT and IRHRPRPY to ISHSPSPY).²

Radiolabeling of Human IL-1-- Recombinant human IL-1 (kind gift of Immunex Corp) was radiolabeled with ¹²⁵I by a modified chloramine-T method as described previously (12, 27). Briefly, 30 ng of IL-1 (1.71nmol) in 10 µl of PBS were incubated with 5mCi (2.0nmol) of sodium ¹²⁵I-iodide (NEN Life Science Products) in 25 µl of 0.5 M sodium phosphate, pH 7, and 30 µl of 1.4 × 10⁴ M chloramine-T (4.2 nmol) for 30 min on ice. The reaction mixture was fractionated and the reaction terminated by rapid filtration on a 1-ml bed volume Biogel P6 column (Bio-Rad) preblocked with bovine serum albumin (1% (w/v) in PBS). Following elution with PBS, fractions 2-4 (100 µl each), containing the labeled protein, were pooled. The labeled protein had a specific activity in the range 3-6 × 10¹⁵ dpm/mmol.

Receptor Binding-- For binding experiments, monolayer cultures plated on fibronectin coated plates or on bare plastic, as described above, were washed twice with fresh medium. Subsequently, binding medium (RPMI 1640 medium with 1% bovine serum albumin, 20 mM HEPES, pH 7.2, and 0.1% sodium azide; 1 ml/10⁵ cells) containing the appropriate concentration of ¹²⁵I-IL-1 was added to the wells. Nonspecific binding was measured using 100-fold higher concentration of unlabeled IL-1. Cultures were incubated at 4 °C for 2 h using a gyrorotary shaker to ensure continuous mixing of the supernatant. The incubation time chosen was based on previous kinetic studies (27). The binding experiments were carried out in binding medium alone or in the presence of heparin (10⁵ M) or chondroitin (10⁵ M) sulfate, used as control, or in the presence of an RGD containing peptide (GRGDSP, 10⁴ M, Life Technologies, Inc.) or a nonspecific, RGE containing, peptide (GRGESP, 10⁴ M, Life Technologies, Inc.) used as a control. In addition, binding was measured in the presence of a blocking anti-IL-1RI monoclonal antibody (M4, 10⁷ M) (35). Association kinetics experiments were performed at four IL-1 concentrations (0.1, 0.75, 1, and 1.5 nM). Following incubation, 60 µl of supernatant was withdrawn to measure free ¹²⁵I-IL-1 concentrations. The monolayers were washed rapidly (5×) with ice-cold binding medium to remove unbound ligand and harvested following incubation at 37 °C for 15 min in trypsin (0.05%)/EDTA (0.02%). Bound ¹²⁵I-IL-1 was measured in a gamma or beta counter. Nonspecific binding data were analyzed by curve fitting using the following equation: Bound (molecules/cell) = A × C, where C is the free IL-1 concentration (M), and A (molecules/cell/M) (determined using one or two IL-1 concentrations) is the slope of a line passing through the origin. Nonspecific binding for all experimental points was subsequently calculated by interpolation and subtracted from the specific binding data. Molecules/cell was calculated using the formula,

(Eq. 1)

where A is Avogadros number, B is the specific activity of the radiolabeled IL-1 (cpm/mol), and C is the number of cells determined in duplicate by hemocytometer.

Values for specific binding (molecules/cell) were analyzed by nonlinear least squares fitting using the equation,

(Eq. 2)

where K (M¹) is the affinity of IL-1 for its receptor, R_o is the total receptor concentration (sites/cell), and C (M) is the IL-1 concentration. Parameter values were estimated by nonlinear least squares fitting of this equation to equilibrium binding data. Association kinetics data were analyzed using a pseudo-first-order treatment as described (27). All calculations were done using MLAB for Macintosh (Civilized Software, Silver Spring, MD). For illustration purposes, in some cases the data were converted to Scatchard format.

Cross-linking and Gel Electrophoresis-- For cross-linking experiments, cells were incubated with radiolabeled ligand and rinsed, as above. Prior to harvesting, cultures were incubated with the amine-directed cross-linking agent BS³ (0.1 mg/ml, Pierce) for 30 min at room temperature (36).

Following harvesting by EDTA (5 mM) and centrifugation (1000 rpm for 5 min at 4 °C), the cell pellet was resuspended in extraction buffer (1% Triton X-100, 2 mM phenylmethylsulfonyl fluoride, 1 µM pepstatin, 1 µM leupeptin; 50 µl/10⁶cells) and incubated for 15 min on ice. Following centrifugation (10,000 rpm for 5 min), supernatant from samples was mixed with an equal volume of 2× sample buffer (0.1 M Tris-HCl, pH 6.8, 1% SDS, 10% glycerol, 0.0025% bromphenol blue, and 0.25% 2-mercaptoethanol) and boiled for 5 min. Extracts were separated by SDS-PAGE (3-12% or 4-12%), using the stacking gel procedure of Laemmli (37). Dried gels were allowed to expose film (X-Omat, Eastman Kodak Co.) with intensifying screens (Hypercassette, Amersham Pharmacia Biotech) for 3-7 days at 70 °C. For quantitation, autoradiograms were scanned with a ScanJet IIcx (Hewlett-Packard, Boise, Idaho) connected to a power Macintosh and analyzed using NIH Image, as above.

Cell Surface Iodination-- Cells plated on fibronectin coated plates, as above (six to ten 10-cm plates at 2 × 10⁶cells/plate) were surface-labeled using the glucose oxidase-lactoperoxidase method (38). Briefly, cultures were incubated (15 min at room temperature) with 2.5 mCi/ml¹²⁵I-NaI (NEN Life Science Products, low pH) in iodination buffer (PBS containing lactoperoxidase (20 µg/ml), glucose oxidase (0.25 µg/ml), and glucose (20.00 mM)). After washing (0.15 M NaI in PBS, 1:1), incubation with unlabeled IL-1 (2 × 10⁹ M for 2 h at 4 °C), and cross-linking (BS³, 0.1 mg/ml), cells were detached (EDTA, 5 mM). The cell pellet, obtained by centrifugation (1000 rpm for 5 min) was resuspended in extraction buffer containing 1% Triton X-100 and protease inhibitors, as above, and incubated for 30 min on ice. The supernatant after centrifugation (13,000 rpm, 5 min) was removed and proteins separated using SDS-PAGE, as above.

Heparinase Digestion-- After cross-linking, samples (50 µl) were digested with heparinase (0.2 units/ml, 6 h at 37 °C) (EC 4.2.2.7) (Sigma) in a Tris-HCl buffer (50 mM, pH 7.0) containing 10 mM CaCl₂, (39) and in the presence of 0.01 mM leupeptin and 0.01 mM pepstatin, with equal amounts of fresh enzyme added at 3 h. Following digestion, samples were analyzed using gel electrophoresis, as above.

Immunoprecipitation-- Cells plated on fibronectin were incubated with radiolabeled IL-1 for 2 h, as described previously. In some experiments, cells were instead biosynthetically labeled using [³⁵S]sulfate (300 µCi/ml, overnight at 37 °C), re-plated on fibronectin following detachment using EDTA, as above, and incubated in the cold (4 °C) with unlabeled IL-1.

Following either protocol, cells were incubated with cross-linking agent, harvested, and extracted, as above.

Following preclearing with secondary antibody (rabbit anti-rat, 290 µg/ml) and protein A-agarose beads (12.5 µl/ml, Santa Cruz Biotechnology Inc.), samples (20 µl/10⁶ cells) were subjected to immunoprecipitation by adding a nonblocking anti-human type I IL-1 receptor rat monoclonal antibody (M8, 5 µg/ml, 1 h at 4 °C) (35). Subsequently, secondary antibody (rabbit anti-rat, Capell) (1:20,290 µg/ml) and prewashed protein A-agarose beads (12.5 µl/ml, Santa Cruz Biotechnology Inc.) were added to each sample, followed by incubation for an additional 1 h at 4 °C.

Samples were spun in a microcentrifuge (10,000 rpm for 1 min at 4 °C) and washed three times with RIPA buffer (150 mM NaCl₂, 50 mM Tris-HCl, pH 7.5, 0.5% deoxycholate, 0.5% Nonidet P-40 and 0.1% SDS) and once with HEPES, (10 mM, pH 7.5). After the final wash, the pellet was resuspended in electrophoresis sample buffer (50 µl/10⁶ cells) and analyzed by SDS-PAGE, as above.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Confocal microscopy of human fibroblasts transfected with a 5'IL-1RI-EGFP 3' receptor construct showed that 24 h after transfection, the receptor fusion protein was located at the cell membrane. Serial observations of single cells replated on fibronectin demonstrated a high degree of accumulation of the GFP-tagged fusion protein at extended processes following cell attachment and spreading (Fig. 1a), in agreement with earlier data on localization of the endogenous receptor protein. Immunocytochemical staining of transfected cells showed co-localization of the receptor fusion protein with the transmembrane linkage protein vinculin at these sites, and at occasional sites along the cell membrane (Fig. 1).

View larger version (61K): [in this window] [in a new window]   Fig. 1.   Co-localization of IL-1 receptors with the transmembrane linkage protein vinculin in attached fibroblasts. Immunocytochemical staining of fibroblasts transfected with an EGFP-IL-1 receptor construct, and following attachment and spreading on fibronectin, shows accumulation of the fusion protein at extended processes (arrowheads) and at occasional sites along the membrane (arrows) (a) and co-localization with endogenous vinculin at these sites (b). Bar, 10 µm.

Binding studies showed that attachment to fibronectin caused an increase in endogenous cell surface receptor levels (Fig. 2a). As reported earlier (11), there was a significant variation in receptor number between experiments (fibronectin, 2.3-11 × 10³ receptors/cell; tissue culture, 1.1-7.3 × 10³ receptors/cell). However, in parallel experiments, cells attached to fibronectin consistently showed a higher level of cell surface binding (4.3 ± 0.85 × 10³ receptors/cell) than cells on bare tissue culture plastic (2.6 ± 0.6 × 10³ receptors/cell), an average increase of 1.8 ± 0.17-fold. Similarly, attachment to collagen and vitronectin resulted in an increase in IL-1 receptor levels, which was, however, less pronounced (1.3-1.5-fold) (data not shown).

View larger version (14K): [in this window] [in a new window]   Fig. 2.   IL-1 receptor levels are affected by fibronectin attachment. a, cells plated on fibronectin () or on bare tissue culture plates () were incubated in the cold (4 °C) for 2 h with various concentrations of radiolabeled IL-1, as indicated, and the level of specific receptor binding was determined. Nonspecific binding, measured in the presence of 100-fold excess unlabeled ligand, has been subtracted from the data, and the continuous curves were calculated from the best fit parameter values. Scatchard analyses reveal a 2-fold reduction in affinity constant in cells attached to fibronectin relative to cells on bare tissue culture plastic. Data represent the mean of four experiments. b, cells were incubated with radiolabeled IL-1 but in the presence or absence of a blocking anti-type I IL-1 receptor antibody, and the level of specific IL-1 receptor binding was measured as in a. Data shown represent mean of two experiments.

Scatchard analyses revealed that fibronectin attachment, in addition, caused a decrease in IL-1 receptor affinity (K_a) for ligand, from 1.0 × 10⁹ (M¹), in cells plated on tissue culture plastic, to 5.6 × 10⁸ (M¹) in fibronectin-attached cultures (Fig. 2a). Furthermore, incubation with a blocking monoclonal antibody to the type I IL-1 receptor resulted in total inhibition of receptor binding under both conditions, showing that the attachment-induced enhancement in specific cell surface IL-1 binding was absolutely dependent on binding to the type I IL-1 receptor (Fig. 2b). Additional analyses of binding kinetics at a range of IL-1 concentrations revealed a 2-fold reduction in association rate constant in cells attached to fibronectin (k_on, 2.42 × 10⁷ M¹ min¹) compared with control (k_on, 4.77 × 10⁷ M¹ min¹) (Fig. 3), demonstrating that the effect on the affinity constant shown in Fig. 2a could be accounted for totally by an effect on the on-rate (k_on).

View larger version (19K): [in this window] [in a new window]   Fig. 3.   Fibronectin attachment causes a reduction in the IL-1/IL-1 receptor association rate constant. a, cells plated on fibronectin (FN) or on bare tissue culture plastic (TC), as above, were incubated with various concentrations of radiolabeled IL-1 (0.1-1.5 nM) for various times, as indicated, and the level of cell surface receptor binding determined. Data shown are for two concentrations and represent the mean of three experiments. b, association rates calculated from data as in a are plotted against IL-1 concentrations. , fibronectin; , tissue culture. The association rate constant for each condition, represented by the slope of the line, is calculated as described under "Experimental Procedures." Data show mean and standard error of the parameter values (using MLAB) for three experiments.

The effects of fibronectin attachment on IL-1 receptor binding could be reversed by the addition of heparin. This effect was concentration-dependent and saturable, resulting in about a 40% reduction in IL-1 receptor binding at 10⁶-10⁵ M soluble heparin (data not shown). Thus, in cells plated on fibronectin, cell surface binding of radiolabeled IL-1, measured in the presence of soluble heparin was similar to that measured in cells on bare plastic, whereas addition of chondroitin sulfate had no effect (Fig. 4a). In comparison, in cells plated on bare plastic, addition of soluble heparin had no effect on receptor binding (Fig. 4b). In contrast, inhibiting integrin aggregation by the use of an RGD peptide (10⁴ M) or adding a control peptide (RGE, 10⁴ M) had no effect on cells plated under either condition (data not shown). The effect of fibronectin attachment on the kinetics of IL-1/IL-1 receptor interaction was also inhibited by addition of heparin (Fig. 5). Thus, the association rate constant for cells attaching to fibronectin (k_on, 2.4 × 10⁷ M¹ min¹) was increased in the presence of heparin (k_on, 5.9 × 10⁷ M¹ min¹), reaching a value similar to that for cells on tissue plastic (k_on, 4.8 × 10⁷ M¹ min¹).

View larger version (16K): [in this window] [in a new window]   Fig. 4.   The fibronectin induced effects on IL-1 receptor levels are inhibited by heparin. Cells were plated on fibronectin or on bare tissue culture plastic in panel a or on bare tissue culture plastic in panel b as above and incubated with radiolabeled IL-1 in the presence or absence of soluble heparin (105 M) or with the same concentration of chondroitin sulfate used as control, and specific IL-1 receptor binding was determined as above. a: , tissue culture; , fibronectin; , fibronectin + heparin; , fibronectin + chondroitin sulfate; b: , tissue culture; , tissue culture + heparin; , tissue culture + chondroitin sulfate. Data show mean of three experiments.

View larger version (24K): [in this window] [in a new window]   Fig. 5.   The fibronectin induced effect on binding kinetics is inhibited by heparin. Cells were plated on tissue culture plates () or on fibronectin (FN) and incubated with various concentrations of radiolabeled IL-1 (0.75-1.5 nM), in the presence of absence of heparin (HEP), as above, and the amount of specific cell surface binding determined after various times of incubation. These data were used to determine the association rate constant as in Fig. 3 and show mean ± S.E. of two experiments.

Chemical cross-linking after incubation with radiolabeled ligand generated a band at 97 kDa, either comprising an IL-1/IL-1 type I receptor complex or resulting from a ligand/accessory protein (AcP) interaction (9, 11). This was present in cells plated either in the presence or absence of fibronectin. In addition, in cells attached to fibronectin, a broad, high molecular mass complex of 300-350 kDa was present that contained between 50 and 70% of the cross-linked IL-1 (Fig. 6). This complex was barely detectable in cells on tissue culture plastic. The presence of 100-fold excess cold ligand resulted in complete blocking of cross-linking of radiolabeled IL-1 to both the 97-kDa band and to the high molecular mass band, demonstrating specificity of both interactions.

View larger version (40K): [in this window] [in a new window]   Fig. 6.   Fibronectin attachment induces formation of a high molecular mass complex. Cells were plated on fibronectin and incubated with radiolabeled IL-1, and IL-1 associated complexes visualized by SDS-PAGE, using gradient gels (4-12%) after cross-linking. H, radiolabeled IL-1 2 nM; H+C, control lane with radiolabeled IL-1 (2 nM) + unlabeled IL-1 (200 nM); FN, fibronectin; TC, bare tissue culture plates.

Enzyme digestion of cell extracts from ¹²⁵I-IL-1 cross-linking experiments showed that incubation with heparinase resulted in loss of the high molecular mass complex, with no effect on the intensity of the ligand/receptor complex at 97 kDa (Fig. 7a). Furthermore, in cells plated on a mutant fibronectin fragment with altered heparin binding sites, no IL-1 cross-linking to the high molecular mass component could be detected. Instead, 100% of the cross-linked IL-1 was present in the 97-kDa band, as in cells on tissue culture plastic (Fig. 7b).

View larger version (44K): [in this window] [in a new window]   Fig. 7.   a, the attachment induced high molecular mass complex contains a heparan sulfate. Cells were plated on fibronectin and incubated with radiolabeled IL-1, as above. Following ligand binding and cross-linking, cell extracts were incubated in the presence or absence of heparinase (0.2 units/ml) for digestion of glycosaminoglycan chains, prior to separation by gel electrophoresis (3-12% SDS-PAGE). b, formation of the high molecular mass component requires interaction with the heparin binding site on fibronectin. Cells were plated on wild type fibronectin (FN), on a mutant fibronectin lacking the heparin binding domain (FN(hep)), or on bare tissue culture plastic (TC). Cells were incubated with radiolabeled IL-1 and complexes cross-linked as above. Separation was done using 3-12% SDS-PAGE.

An interaction between the IL-1 receptor and the high molecular mass component was demonstrated by immunoprecipitation, following IL-1 binding and cross-linking. Thus, a nonblocking monoclonal antibody to the type I IL-1 receptor immunoprecipitated both the 97-kDa and the 300-350-kDa complexes from fibronectin plated cells to which IL-1 had been cross-linked (Fig. 8a). This procedure, after cell surface iodination and cross-linking, followed by heparinase treatment, revealed a core protein of the recruited proteoglycan of 80-90 kDa (Fig. 8b). Furthermore, immunoprecipitation following sulfate labeling of the glycosaminoglycan chains after incubation with or without IL-1 showed that the appearance of the high molecular mass band was dependent on ligand binding to the type I receptor (Fig. 8c).

View larger version (29K): [in this window] [in a new window]   Fig. 8.   a, the fibronectin-induced high molecular mass complex interacts with the IL-1 receptor. Cells were plated on fibronectin and incubated with radiolabeled IL-1 as above. Following cross-linking, the extract was immunoprecipitated with an anti-IL-1 receptor antibody and subsequently subjected to 3-12% SDS-PAGE. The negative control shown (2nd Ab) is an immunoprecipitation carried out in the presence of the second antibody and beads alone. b, the heparan sulfate contains a core protein of 80-90 kDa. Immunoprecipitation with an anti-IL-1 receptor monoclonal antibody (MoAb) was carried out following cell surface iodination of cells plated on fibronectin and incubated with unlabeled IL-1 (1 nM). Extracted samples were incubated with (+Hep) or without (Hep) heparinase (0.2 unit/ml) and analyzed by 3-12% SDS-PAGE. c, IL-1 receptor/heparan sulfate interaction is stabilized through ligand binding. Cells were plated on fibronectin, biosynthetically labeled using [35S]sulfate (overnight, 37 °C), replated on fibronectin, and incubated with or without unlabeled IL-1 as indicated, followed by cross-linking. Cell extracts were immunoprecipitated using an anti IL-1 receptor antibody and separated by gel electrophoresis (3-12% SDS-PAGE).

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Glycosylated proteins have been shown to play an important role in cytokine and growth factor regulation of biological responses (40, 41). The effects involve control of signaling (2, 42), as well as receptor function (1, 43-45). Transfection experiments showed the IL-1 receptor protein in adherent cells to be located at sites of focal adhesion, in agreement with earlier reports using radiolabeled ligand (27). Fibronectin attachment through these sites induced formation of a high molecular mass IL-1 receptor complex containing cell surface heparan sulfate, resulting in effects on IL-1 receptor expression and affinity. Induction of these effects was mediated through attachment via the heparin binding domain in fibronectin and stabilized by IL-1/receptor interaction.

The increase in type I IL-1 receptor expression induced by fibronectin may reflect an increase in the rate of translocation of receptor protein to the cell surface. It is, however, known that translocation of receptor protein is rapid and that over 95% is normally expressed at the cell surface (46). Alternatively, the increase in steady state surface expression could result from retardation in the basal internalization rate, by stabilization in focal adhesions at the cell surface. Such a mechanism, involving masking or immobilization of the receptor protein, is suggested by the slower association rate of the ligand. In addition, this is consistent with the recruitment of the additional high molecular mass component to the IL-1 receptor complex. Finally, the lack of effect using the RGD peptide demonstrated that integrin mediated signaling was not involved and supports the notion that the changes were induced as a direct consequence of alterations in the ligand or in the receptor complex.

Involvement of heparan sulfate proteoglycan in matrix regulation of IL-1 receptor function was suggested by inhibition of the fibronectin induced effects by soluble heparin and confirmed by digestion of the high molecular mass complex with heparinase. In contrast to heparin regulation of fibroblast growth factor-receptor binding (1), in which the ligand binds the proteoglycan with detectable affinity, no interaction occurs initially between IL-1 and the proteoglycan, as suggested by binding studies on cells attached to bare plastic. The effect involves a cell surface component in close proximity to the IL-1 receptor as demonstrated by cell surface labeling, cross-linking and immunoprecipitation. Furthermore, the heparan sulfate is recruited secondary to IL-1/IL-1 receptor binding, as demonstrated by blocking experiments using a monoclonal antibody or excess cold ligand. The effect involves an increase in a single class of IL-1 binding sites, as judged by affinity constant, again indicating that the IL-1 affinity for the heparan sulfate is low and nondetectable in conventional binding assays. Interfering with IL-1/IL-1 receptor binding required a 10,000-fold higher concentration of the glycosaminoglycan, suggesting an interaction with a K_a in the range of 10⁵ M similar to that of integrin/matrix interaction and about 100-fold lower than that of heparan-fibroblast growth factor interaction (1).

The requirement for ligand/IL-1 receptor binding, prior to association of the heparan sulfate suggests a mechanism similar to that in the LIF/oncostatin M system, in which recruitment of gp130 to the receptor complex follows LIF binding (2, 42). Digestion experiments suggested the ligand/heparan sulfate interaction to be due to protein/glycosaminoglycan binding, the most common type of growth factor/proteoglycan interaction. Similar to the heparin-glycosminoglycan-fibroblast growth factor interaction (47-49), it is likely also dependent on sulfation (50) and could involve a conformational change of the IL-1 protein following interaction with the glycosaminoglycan (51). Effects of heparan sulfate on the receptor protein are likely indirect, because the only species appearing following sulfate labeling and cross-linking was of the same size as that containing the heparan sulfate and the ligand, observed using radiolabeled IL-1. Thus, the simplest model suggested by this data is that following binding to fibronectin, cell surface heparan sulfate is recruited to the IL-1 receptor complex by IL-1 and that IL-1 functions as a bridge between the receptor and heparan sulfate. Furthermore, the data show that effects of the heparan sulfate binding are induced indirectly through the ligand and that no direct interaction occurs between the two membrane components during complex formation.

The role of the proteoglycan in this system may be similar to that of -glycan, the nonsignaling type III TGF- receptor, affecting biological responses in an indirect manner by regulating binding affinity of the signaling type I receptor (43, 52). The relative level of IL-1 bound to the heparan sulfate was estimated as 50-70%, correlating well with the fraction of receptors located at focal adhesions (27) and supporting the notion that this interaction is an adhesion regulated event. The heparan sulfate in this system may thus function like the accessory protein (AcP) (9), although specifically responding to cell attachment. This is further supported by the dependence on heparan sulfate/fibronectin interaction, as demonstrated by the negligible levels of the high molecular mass complex observed in cells on bare plastic as well as using the mutant fibronectin lacking the heparin binding site.

These types of alterations in IL-1 receptor/ligand interaction correlate well with attachment-mediated effects on IL-1 signal transduction (16, 30). For example, in the presence of an increasing concentration of soluble heparin, NF-B activation by IL-1 in cells plated on fibronectin is progressively inhibited. In addition, attachment-mediated effects could be a direct consequence of matrix or IL-1 binding the cell surface heparan sulfate. Cell surface HSPGs, such as members of the syndecan family (53-55), can induce signal transduction through effects on PKC dependent pathways (56-59). Both mechanisms of inducing alterations in signaling could be in part responsible for matrix regulation of IL-1 biological responses (31). Regulation of heparan sulfate proteoglycans by cytokines (60-62) and their simultaneous influence on cytokine and growth factor action (63-67) suggest that this interaction could mediate feedback inhibition of IL-1 activity. Thus, release of soluble heparin at sites of injury could inhibit IL-1 release of matrix degrading enzymes and favor matrix regeneration during later stages of wound healing in inflammation.

In summary, our data show an effect of fibronectin attachment on IL-1 receptor function, involving a cell surface heparan sulfate. The effect appears to be proximally mediated by the recruitment of an HSPG to the receptor complex following IL-1/IL-1 receptor binding in the presence of fibronectin. The presence of the attachment-induced component, recruited to the IL-1 receptor complex, correlates with structural effects on IL-1 signaling and gene regulation. Ongoing experiments aim to identify this cell surface heparan sulfate and to determine how the alterations induced in receptor binding affect recruitment of other components of the complex, such as the transmembrane accessory protein (AcP) component (9, 68); intracellular components TRAF 6 (23), IIp 1 (69), and MyD 88 (24, 25, 26); and kinases of the IRAK/Pelle family (7, 24).

	ACKNOWLEDGEMENTS

IL-1 and IL-1 were a generous gift from Immunex Corp. (Seattle, WA). Autoradiographic analysis was carried out in part by the PhosphorImager Facility of the Markey Molecular Medicine Center at the University of Washington, Seattle. The confocal microscopy analysis was carried out using the facility in the Laboratory of Molecular Immunology, University of Sheffield.

	FOOTNOTES

^* This work was supported by Grants DE-10363 and DE-11251 from the National Institutes of Health, Grant Q0502 from the Arthritis and Rheumatism Council, Grant PG97 112 from the British Heart Foundation (to E. E. Q.), a grant from the Wellcome Trust (to M. J. H.), and grants from the Wellcome Trust and the Medical Research Council (to S. K. D.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

To whom correspondence should be addressed. Tel.: 44-114-271-3181; Fax: 44-114-271-3846; E-mail: E.Qwarnstrom@Sheffield.ac.uk.

² J. A. Askari and M. J. Humphries, manuscript in preparation.

	ABBREVIATIONS

The abbreviations used are: IL-1RI, interleukin type I receptor; IL, interleukin; PBS, phosphate-buffered saline; PAGE, polyacrylamide gel electrophoresis; EGFP, enhanced green fluorescent protein.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES