Lectin-like Characteristics of Recombinant Human Interleukin-1beta Recognizing Glycans of the Glycosylphosphatidylinositol Anchor*

(Received for publication, January 14, 1997, and in revised form, February 10, 1997)

Keiko Fukushima , Sayuri Hara-Kuge , Takashi Ohkura , Akira Seko , Hiroko Ideo , Toshiyuki Inazu Dagger and Katsuko Yamashita §

From the Department of Biochemistry, Sasaki Institute, Kanda-Surugadai, Chiyoda-ku, Tokyo 101 and the Dagger  Noguchi Institute, Kaga, Itabashi-ku, Tokyo 173, Japan

ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
FOOTNOTES
REFERENCES


ABSTRACT

We found that 35S-labeled recombinant human interleukin-1beta (rhIL-1beta ) binds phosphatidylinositol-specific phospholipase C-treated human placental alkaline phosphatase, phosphatidylinositol-specific phospholipase C-treated trypanosome surface variant glycoproteins, and urinary uromodulin immobilized on plates or immobilized on CNBr-activated Sepharose 4B. The interaction between rhIL-1beta and these glycoproteins was lectin-like, since it was inhibited in the presence of specific saccharides, i.e. mannose 6-phosphate or synthetic Ac-NH·CH2·CH2·PO4-right-arrow6Manalpha 1right-arrow(±2Manalpha 1right-arrow±6Manalpha 1right-arrow)propyl at about 1 µM. On the other hand, a wide variety of compounds including biantennary sugar chains derived from these glycoproteins as well as ethanolamine phosphate, inositol phosphate, mannose 6-sulfate, mannose 1-phosphate, glucose 6-phosphate, and mannitol 6-phosphate did not show any inhibitory effect at concentrations up to 1 mM. These results indicate that rhIL-1beta interacts with these glycoproteins via the mannose 6-phosphate diester of glycans on the glycosylphosphatidylinositol (GPI) anchor. Furthermore, when monolayers of polarized Madin-Darby canine kidney cells on polycarbonate filter membranes were incubated with 35S-rhIL-1beta in either the apical or basolateral chamber, 35S-interleukin-1beta was found to bind specifically to the apical membranes with a Ka value of 4.6 × 107 M-1, and the specific interaction was inhibited by 1 µM mannose 6-phosphate. Since the mannose 6-phosphate diester moiety exists only in the GPI glycans on plasma membranes, it was evident that interleukin-1beta can directly interact with the mannose 6-phosphate diester component of the intact glycan of GPI anchors on plasma membranes.


INTRODUCTION

Until now, over 50 types of cytokines have been reported that mediate the regulatory network established between lymphoid cells, hematopoietic cells, and endothelial cells that control their differentiation and proliferation (1). In this cytokine network, a single cytokine can act as both a positive signal and a negative one, depending on the type of target cells. In contrast, in some instances multiple cytokines act on a single cell and show the same biological effects. The response of target cells to a given cytokine is determined by the expression of the cytokine receptor and/or the nature of the link between the receptor and the signal transduction pathways of the target cells. We have been interested in the lectin-like character of cytokines in relation to their role as signal transducers or effectors. On the basis of their lectin-like character, various cytokines have been categorized into several groups. Growth factors such as granulocyte-macrophage colony-stimulating factor (2), heparin-binding growth factor (3), basic fibroblast growth factor (4, 5), and midkine (6) bind to heparan sulfate, whereas tumor necrosis factor-alpha (TNF-alpha )1 (7), lymphotoxin (8), interleukin-1alpha and -1beta (7, 9), interleukin-2 (10), and interleukin-3 (11) recognize N,N'-diacetylchitobiose and/or some oligomannosyl residues, although the precise carbohydrate binding specificities and target glycoproteins have not yet been determined. To elucidate the physiological role of the lectin-like interaction between cytokines and target molecules, the carbohydrate binding specificities of cytokines should first be determined. We describe in this paper the lectin-like interaction between recombinant human IL-1beta (rhIL-1beta ) and target glycoproteins by measuring the binding of 35S-labeled rhIL-1beta to various glycoproteins immobilized on Sepharose 4B, on plates coated with phosphatidylinositol-specific phospholipase C (PI-PLC)-treated human placental alkaline phosphatase (ALP), and on polarized Madin-Darby canine kidney (MDCK) cells and also by assaying the inhibition of this interaction by various haptenic compounds.


EXPERIMENTAL PROCEDURES

Materials

Human transferrin, bovine ribonuclease B, mannose 6-phosphate, mannose 1-phosphate, glucose 6-phosphate, myo-inositol 1-phosphate, mannose 6-sulfate, and ethanolamine phosphate were purchased from Sigma. [35S]Methionine (10.27 Ci/mmol) was obtained from ICN Pharmaceutical Inc. Arthrobacter ureafaciens sialidase and Bacillus thuringiensis PI-PLC were purchased from Nacalai Tesque (Kyoto) and Funakoshi Inc. (Tokyo), respectively. Mannitol 6-phosphate was prepared from mannose 6-phosphate by reduction with NaBH4. H2N·CH2·CH2·PO4-right-arrow6(Manalpha 1right-arrow2)mannitol was kindly provided by Dr. S. Iwahara (Kagawa University, Kagawa, Japan) and H2N·CH2·CH2· PO4-right-arrow6mannitol was prepared from H2N·CH2·CH2·PO4-right-arrow6(Manalpha 1right-arrow2)mannitol by Aspergillus saitoi Manalpha 1right-arrow2-specific alpha -mannosidase digestion (12). Galbeta 1right-arrow4GlcNAcbeta 1right-arrow2Manalpha 1right-arrow6 (Galbeta 1right-arrow4GlcNAcbeta 1right-arrow 2Manalpha 1right-arrow3)Manbeta 1right-arrow4GlcNAc (abbreviated as Gal2·GlcNAc2·Man3· GlcNAc) was purified from the urine of patients with GM1-gangliosidosis (13). ALP was a generous gift from Dr. M. Ikehara (Fukuoka University, Fukuoka, Japan). The ALP was purified from human placenta by the method of Ogata et al. (14). MDCK strain II cells were kindly supplied by Dr. M. Tashiro (National Institute of Health, Tokyo, Japan).

Synthesis of Phosphate Derivatives

Ac-NH·CH2·CH2·PO4-right-arrow 6Manalpha 1right-arrowOCH2·CH2·CH3, Ac-NH·CH2·CH2·PO4-right-arrow6Manalpha 1right-arrow2Manalpha 1right-arrow OCH2·CH2·CH3, and Ac-NH·CH2·CH2·PO4-right-arrow6Manalpha 1right-arrow2Manalpha 1right-arrow 6Manalpha 1right-arrowOCH2·CH2·CH3 were synthesized according to the hydrogen phosphate method as described by Pannecoucke et al. (15) using van Boom reagent (16) from N-acetylethanolamine and the corresponding suitable benzylated allyl mannopyranosides that were prepared by the dimethylphosphinothioate method (17).

Preparation of Uromodulin

Uromodulin was prepared from pooled urine (1 liter) from a pregnant donor (in the 12th week of pregnancy) as follows (18). 0.58 mol of NaCl was added to the urine, and it was stirred at 4 °C overnight. The precipitate formed was collected by centrifugation and dialyzed against distilled water. After centrifugation, the supernatant was freeze-dried and dissolved in phosphate-buffered saline (PBS).

Preparation of Trypanosome Variant Surface Glycoproteins

Trypanosomes (Trypanosoma brucei gambiense) isolated from the blood of an infected mouse according to the method described by Cross (19) were kindly provided by Dr. Y. Tachibana (Tokai University, Japan). A 100-µl sample of packed trypanosomes was digested with B. thuringiensis PI-PLC (0.5 units/ml in Tris-buffered saline, pH 7.4, 37 °C, 5 h), and the solubilized variant surface glycoproteins (VSG) were purified by Superose 12 (Pharmacia Biotech Inc.) gel permeation chromatography. Upon analysis by SDS-polyacrylamide gel electrophoresis according to the method of Laemmli (20) followed by silver staining, the purified VSG appeared as a broad single band with a molecular mass of about 60 kDa.

Preparation of rhIL-1beta

cDNA encoding human IL-1beta (R&D Systems Europe Ltd., Abingdon, United Kingdom) was used to produce rhIL-1beta in Escherichia coli. Plasmid pET3a (Novagen, Inc., Madison, WI) was used as the T7 expression plasmid (21). A NdeI-BamHI fragment corresponding to the synthetic human IL-1beta gene was inserted between the NdeI and BamHI sites of pET3a to produce the expression plasmid. The rhIL-1beta gene was expressed in E. coli strain BL21(DE3) under the control of the T7 promoter. The recombinant protein was released from the cells by osmotic shock, concentrated with an Amicon PM10 Diaflo membrane, and applied to a Sephacryl S-200 column. Fractions containing rhIL-1beta were pooled and stored at -20 °C until use.

Preparation of 35S-rhIL-1beta

The cloned cDNA encoding IL-1beta (R&D Systems Europe Ltd., Abingdon, UK) inserted downstream from the T7 RNA polymerase promoter in plasmid pET3a was used as a template for in vitro transcription and translation in the Single Tube Protein System 2, T7 (Novagen, Inc.) in the presence of [35S]methionine. The translation products were separated from free [35S]methionine using a PD-10 column (Pharmacia). One reaction using 2.5 µg of plasmid DNA templates, 200 µl of lysate, and 200 µCi of [35S]methionine provided approximately 500 fmol of 35S-rhIL-1beta (0.9 × 107 dpm/pmol).

Solid-phase Binding Assay

The binding of 35S-labeled rhIL-1beta to various glycoproteins was measured by a solid-phase binding assay. Enzyme-linked immunosorbent assay plates (Corning, Inc., Corning, NY) were coated with glycoproteins at 10 µg/ml in 20 mM carbonate buffer, pH 9.6, at 4 °C overnight. The plates were washed with 0.05% Tween 20 in PBS, pH 7.3, blocked with PBS, 0.05% Tween 20, 1% BSA and then treated with 1 × 105 dpm of 35S-rhIL-1beta in PBS, 0.05% Tween 20, 0.1% BSA at 37 °C for 2 h. After washing with 0.05% Tween 20 in PBS, the bound 35S-rhIL-1beta was released by treatment with 100 µl of 1 M acetic acid, and the radioactivity was measured by means of a liquid scintillation counter.

Preparation of Sepharose 4B-immobilized Glycoproteins

ALP (1.2 mg) was dissolved in 1 ml of coupling buffer (0.1 M NaHCO3 (pH 8.3) containing 0.5 M NaCl). 1 ml of CNBr-activated Sepharose 4B (Pharmacia) in a column (inner diameter 0.7 cm × 5 cm long) was washed with 1 mM HCl (20 ml), and the HCl was replaced with the coupling buffer. The ALP solution was quickly added to the CNBr-activated Sepharose 4B resin. After gentle mixing at 4 °C for 18 h, the gel was washed with 5 gel volumes of coupling buffer. The gel was shaken in an equal volume of 0.2 M glycine, pH 8.0, at room temperature for 2 h and washed with 0.1 M acetate buffer containing 1 M NaCl, pH 4.0, and with coupling buffer (three times each). The amount of ALP bound to the Sepharose 4B (based on the amount of ALP remaining in the coupling reaction filtrate) was estimated to be approximately 1 mg/ml gel. Bovine ribonuclease B-Sepharose (10 mg/ml gel), trypanosome VSG-Sepharose (1 mg/ml gel), uromodulin-Sepharose (1 mg/ml gel), and human transferrin-Sepharose (10 mg/ml gel) were prepared according to the same procedures as those employed for ALP-Sepharose.

Fractionation of 35S-rhIL-1beta on Sepharose 4B Column-immobilized Glycoproteins

5000 dpm of 35S-rhIL-1beta was applied to a column (inner diameter, 0.7 cm × 5 cm long) of ALP-, VSG-, uromodulin-, transferrin-, or ribonuclease B-Sepharose (1 ml), which was equilibrated with PBS containing 0.1% BSA (PBS-BSA) at 4 °C. After standing for 30 min, the respective columns were washed with 5 column volumes of PBS-BSA and the bound 35S-rhIL-1beta was eluted with PBS-BSA containing 1 mM mannose 6-phosphate.

Cell Culture

MDCK strain II cells were maintained in minimal essential medium (MEM) supplemented with 10% fetal bovine serum. For binding experiments, MDCK cells that were plated at confluent density (approximately 3.5 × 105 cells) in 6.5-mm Transwell filter chambers (Costar) and cultured for 4 days after plating were used. The polarity of cells was assessed by polarized uptake of [35S]methionine from the basolateral side (>10:1, basolateral:apical) (22).

Binding of 35S-rhIL-1beta to MDCK Cell Monolayers

The polarized monolayers on Transwell filters were washed with ice-cold MEM (200 µl), the same medium was added to both the apical (0.1-ml) and basolateral (0.5-ml) chambers, and the cells were incubated for 1 h at 4 °C. The medium in both chambers was changed to media with or without 200 µg/ml rhIL-1beta , and the cells were incubated for another hour at 4 °C. For assay of apical binding, the medium in the apical compartment was replaced with 50 µl of medium containing 0.2 µCi/ml 35S-rhIL-1beta with or without 200 µg/ml rhIL-1beta , and the Transwell filters were placed onto 50 µl of MEM medium on a piece of parafilm. For assay of basolateral binding, the non-radioactive medium and the radioactive medium were added to opposite chambers, and the cells were then incubated for the indicated times at 4 °C. The binding assays were terminated by rapid washing of each compartment with ice-cold PBS four times. The cells were lysed in 100 µl of ice-cold Nonidet P-40 lysis buffer (50 mM Tris-Hcl, pH 8, containing 10 mM EDTA, 1% Nonidet P-40, and 0.1% SDS) and removed by gentle scraping. The lysates were centrifuged in a microcentrifuge at 10,000 × g for 10 min to precipitate debris and intact nuclei. Proteins in the supernatants were precipitated with 10% trichloroacetic acid at 4 °C for 30 min, and the samples were centrifuged. The pellets were washed with ethyl ether, and radioactivity was counted using a scintillation counter.


RESULTS

Binding of 35S-rhIL-1beta to Glycoproteins Immobilized on Columns

It has been reported that rhIL-1beta exhibits lectin-like characteristics (7, 9). As the first step to examine the lectin-like characteristics of IL-1beta , the binding of 35S-rhIL-1beta to various glycoproteins immobilized on Sepharose 4B columns was investigated. As shown in Fig. 1, 35S-rhIL-1beta bound to PI-PLC-treated ALP-Sepharose, PI-PLC-treated VSG-Sepharose, or uromodulin-Sepharose (1 mg/ml gel) columns, and the bound 35S-rhIL-1beta was eluted with PBS-BSA containing 1 mM mannose 6-phosphate, which is a constituent of the glycan portion of the glycosylphosphatidylinositol (GPI) anchor (Fig. 1, A, B, and C). In contrast, rhIL-1beta did not bind to human transferrin-Sepharose or bovine ribonuclease B-Sepharose (10 mg/ml gel) columns, where the glycan moiety linked to the protein consists of sialylated biantennary sugar chains (23) and high mannose-type sugar chains, respectively (24) (Fig. 1, D and E). The glycoproteins interacting with rhIL-1beta not only have asparagine-linked sugar chains but also glycans of the GPI anchor at the carboxyl-terminal end. The structures of the glycan portions of the GPI anchor of ALP (25) and trypanosome VSG (26) have already been determined, and their side chain structures differ from each other as shown below.
<AR><R><C><UP>           O</UP></C></R><R><C><UP>           ∥</UP></C></R><R><C><UP>ALP</UP>:<UP>Protein</UP> · <UP>CNH</UP> · <UP>CH</UP><SUB>2</SUB> · <UP>CH</UP><SUB>2</SUB> · <UP>PO</UP><SUP><UP>−</UP></SUP><SUB>4</SUB> → 6(<UP>±NH</UP><SUB>2</SUB> · <UP>CH</UP><SUB>2</SUB> · <UP>CH</UP><SUB>2</SUB> · <UP>PO</UP><SUP><UP>−</UP></SUP><SUB>4</SUB>→)</C></R></AR>
(<UP>Man</UP>&agr;1 → 2<UP>Man</UP>&agr;1 → 6<UP>Man</UP>&agr;1 → 4 <UP>GlcNH</UP><SUB>2</SUB>) → <UP>inositol</UP> · <UP>PO</UP><SUP> − </SUP><SUB>4</SUB>
<AR><R><C><UP>           O</UP></C></R><R><C><UP>           ∥</UP></C></R><R><C><UP>VSG</UP>:<UP>Protein</UP> · <UP>CNH</UP> · <UP>CH</UP><SUB>2</SUB> · <UP>CH</UP><SUB>2</SUB> · <UP>PO</UP><SUP> − </SUP><SUB>4</SUB> → 6<UP>Man</UP>&agr;1 → 2<UP>Man</UP>&agr;1 → 6(<UP>±Gal</UP>&agr;1</C></R></AR>
→ 2<UP>Gal</UP>&agr;1 → 6(<UP>Gal</UP>&agr;1 → 2)<UP>Gal</UP>&agr;1 → 3)<UP>Man</UP>&agr;1 → 4<UP>GlcNH</UP><SUB>2</SUB> → <UP>inositol</UP> · <UP>PO</UP><SUP> − </SUP><SUB>4</SUB>
<UP>S<SC>tructure</SC> I</UP>


Fig. 1. Binding of 35S-rhIL-1beta to various glycoprotein-Sepharose 4B columns. To each column (0.7 cm inner diameter × 5 cm long) 5 × 103 dpm of 35S-rhIL-1beta was applied at 4 °C. The column was eluted with PBS-BSA and then with PBS-BSA containing 1 mM mannose 6-phosphate from the position indicated by the black arrows. A, PI-PLC-treated placental ALP-Sepharose (1 mg/ml gel); B, PI-PLC-treated trypanosome VSG-Sepharose (1 mg/ml gel); C, uromodulin-Sepharose (1 mg/ml gel); D, human transferrin-Sepharose (10 mg/ml gel); E, bovine RNase B-Sepharose (10 mg/ml gel).
[View Larger Version of this Image (17K GIF file)]


Since the 1-O-alkyl-2-O-acylglycerol moiety of the GPI anchor is removed by PI-PLC digestion, it was evident that rhIL-1beta recognizes the common structure, NH2·CH2·CH2·PO4-right-arrow 6Manalpha 1right-arrow2Manalpha 1right-arrow6Manalpha 1right-arrow4GlcNH2, of the glycans in the GPI anchor.

Conditions for Solid-phase Binding Assay of 35S-rhIL-1beta

In binding assays using plates coated with each of the five glycoproteins (10 µg/ml), 35S-rhIL-1beta bound to PI-PLC-treated ALP, PI-PLC-treated VSG, and uromodulin in a dose-dependent manner (data not shown). For determination of the character of the interaction between rhIL-1beta and these glycoproteins, an assay of inhibition of solid-phase binding by specific haptens was employed. PI-PLC-treated human placental ALP was used as the glycoprotein to be immobilized on plates because the structures of both the N-linked sugar chain (27) and the glycan portion of the GPI anchor (25) had been previously determined and are rather simple in comparison with other glycoproteins. We investigated the inhibitory effects of the biantennary sugar chain and various simple components of the glycan portion of the GPI anchor on the interaction between 35S-rhIL-1beta and ALP-coated plates. At first, we investigated the conditions for solid-phase binding of 35S-rhIL-1beta to ALP-coated plates. As for the time dependence, the linearity of the binding activity was maintained for at least 3 h (Fig. 2A, bullet ), and the binding activity was inhibited in the presence of excess amounts of unlabeled rhIL-1beta (Fig. 2A, open circle ). These results indicated that the binding of 35S-rhIL-1beta to ALP is specific and is competitive with unlabeled rhIL-1beta . The binding of 35S-rhIL-1beta was dependent on the concentration of both ALP (Fig. 2B) and 35S-rhIL-1beta (Fig. 2C, bullet ), and the concentration dependence was linear at least up to 20 µg/ml (ALP) and 1 × 106 dpm/ml (35S-rhIL-1beta ), respectively. The reactivity did not change in the presence of 1 mM EDTA, 10 mM HEPES buffer, pH 7.3, 0.1% BSA (Fig. 2C, open circle ) or 1 mM CaCl2, 10 mM HEPES buffer, pH 7.3, 0.1% BSA (Fig. 2C, black-triangle) showing that the lectin-like interaction between rhIL-1beta and ALP at least does not require calcium ions. On the basis of these results, the following inhibitory effects of various haptens were investigated in the proportional range of the protein concentration dependence and time dependence.


Fig. 2. Binding of 35S-rhIL-1beta to ALP-coated plates. Plates were coated with ALP as described under "Experimental Procedures," and the amount of bound 35S-rhIL-1beta was measured using a liquid scintillation counter after being released with 1 M AcOH. A, time course. Plates coated with 10 µg/ml ALP were incubated with 1 × 105 dpm of 35S-rhIL-1beta alone (bullet ) or with 1 × 105 dpm of 35S-rhIL-1beta and 10 µg/ml unlabeled rhIL-1beta (open circle ) for the indicated times at 37 °C; B, ALP concentration dependence. Plates coated with various concentrations of ALP (0-20 µg/ml) were incubated with 1 × 105 dpm of 35S-rhIL-1beta for 2 h at 37 °C; C, 35S-rhIL-1beta concentration dependence. Plates coated with 10 µg/ml ALP were incubated for 2 h with various concentrations of 35S-rhIL-1beta (up to 5 × 106 dpm/ml) in PBS-BSA (bullet ), in 10 mM HEPES, 0.1%BSA, pH 7.3, containing 1 mM EDTA (open circle ), or in 10 mM HEPES, 0.1% BSA, pH 7.3 containing 1 mM CaCl2 (black-triangle).
[View Larger Version of this Image (17K GIF file)]


Inhibitory Effects of Various Haptens on the Lectin-like Activity of rhIL-1beta

The inhibitory effects of various saccharides listed in Table I on the lectin-like interaction between IL-1beta and ALP-coated plates were examined, and the inhibition curves obtained are shown in Fig. 3. The biantennary sugar chain of ALP, which is an asialo-N-linked sugar chain, did not show any inhibitory effect at concentrations up to 10 mM (Fig. 3A, Table I). Since the same results were obtained even when the plates were coated with sialidase-treated ALP (data not shown), the result is not due to a sialylated biantennary sugar chain of ALP. In contrast, mannose 6-phosphate, which is a component of the glycan of the GPI anchor, was an effective inhibitor (Fig. 3B). However, other components of this glycan such as ethanolamine phosphate and inositol phosphate did not show any inhibitory effect at 1 mM (see Table I). To elucidate whether mannose 6-phosphate-diester derivatives show inhibitory effects, three synthetic phosphodiester derivatives as shown in Table I were studied. Ac-NH·CH2·CH2·PO4-right-arrow 6Manalpha 1right-arrowpropyl, Ac-NH·CH2·CH2·PO4-right-arrow6Manalpha 1right-arrow2Manalpha 1right-arrow propyl, and Ac-NH·CH2·CH2·PO4-right-arrow6Manalpha 1right-arrow2Manalpha 1right-arrow6Manalpha 1right-arrowpropyl showed the same strong inhibitory effects as mannose 6-phosphate (Fig. 3, C, D, and E). Mannose 6-sulfate, mannose 1-phosphate, and glucose 6-phosphate did not show any inhibitory effect at concentrations up to 1 mM (Table I), indicating that mannose substituted with phosphate at the C-6 position is necessary for lectin-like binding. Furthermore, mannitol-6-phosphate and NH2CH2·CH2·PO4-right-arrow6mannitol, which are reduced forms of mannose 6-phosphate and the phosphodiester derivative, did not show any inhibitory effect at concentrations up to 1 mM. These results showed that a mannopyranoside substituted with phosphate at the C-6 position represents an important locus for the lectin-like interaction between IL-1beta and ALP, and IL-1beta recognizes the mannose 6-phosphate diester moieties of the glycan of GPI-anchored glycoproteins.

Table I.

Inhibition of 35S-rhIL-1beta binding to plates coated with ALP by various compounds

Binding of 35S-rhIL-1beta to ALP was measured by a solid-phase binding assay as described under "Experimental Procedures."


Compounds Concentration for 50% inhibition

H2N·CH2·CH2·HPO4- No inhibition at 1 mM
HPO4- right-arrow inositol No inhibition at 1 mM
HPO4- right-arrow 6mannose 0.001 mM
HPO4- right-arrow 6glucose No inhibition at 1 mM
HPO4- right-arrow 1mannose No inhibition at 1 mM
HSO4- right-arrow 6mannose No inhibition at 1 mM
HPO4- right-arrow 6mannitol No inhibition at 1 mM
H2N·CH2·CH2·PO4- right-arrow 6mannitol No inhibition at 1 mM
AcNH·CH2·CH2·PO4- right-arrow 6Manalpha 1 right-arrow OCH2·CH2·CH3 0.001 mM
Ac·NH·CH2·CH2·PO4- right-arrow 6Manalpha 1 right-arrow 2Manalpha 1 right-arrow OCH2·CH2·CH3 0.001 mM
AcNH·CH2·CH2·PO4- right-arrow 6Manalpha 1 right-arrow 2Manalpha 1 right-arrow 6Manalpha 1 right-arrow OCH2·CH2·CH3 0.001 mM
Galbeta 1 right-arrow 4GlcNAcbeta 1 right-arrow 2Manalpha 1 down-right-arrow  
                                                                                     6
                                                                                         Manbeta 1 right-arrow 4GlcNAc No inhibition at 10 mM
                                                                                      3
Galbeta 1 right-arrow 4GlcNAcbeta 1 right-arrow 2Manalpha 1 north-east-arrow


Fig. 3. Inhibition curves obtained for the binding of 35S-rhIL-1beta to ALP-coated plates in the presence of various saccharides. A, Gal2·GlcNAc2·Man3·GlcNAc; B, mannose 6-phosphate; C, Ac-NH·CH2·CH2·PO4-right-arrow6Manalpha 1right-arrowpropyl; D, Ac-NH·CH2·CH2·PO4-right-arrow 6Manalpha 1right-arrow2Manalpha 1right-arrowpropyl; E, Ac-NH·CH2·CH2·PO4-right-arrow6Manalpha 1right-arrow 2Manalpha 1right-arrow6Manalpha 1right-arrowpropyl.
[View Larger Version of this Image (17K GIF file)]


Binding of 35S-rhIL-1beta to MDCK Cell Monolayers

All of the results so far described were obtained using PI-PLC-treated glycoproteins. To determine whether intact GPI-anchored glycoproteins on plasma membranes interact with 35S-rhIL-1beta , the following experiments were performed. Because it has been reported that GPI-anchored glycoproteins are distributed on the apical surface of the plasma membrane of polarized MDCK cells (28), and no IL-1 receptor on the plasma membrane of MDCK cells has been reported, we examined the lectin-like interaction between rhIL-1beta and cell surface glycoproteins using polarized MDCK cells as a model. Polarized MDCK cells on polycarbonate filter membranes were incubated with 0.2 pmol/ml 35S-rhIL-1beta in either the apical or basolateral chamber. The time course of binding at 4 °C is shown in Fig. 4. The amount of 35S-rhIL-1beta bound to the cells (Fig. 4, black-square) was reduced by the addition of excess unlabeled rhIL-1beta (Fig. 4, open circle ), suggesting that 35S-rhIL-1beta and unlabeled rhIL-1beta compete for a limited number of specific binding sites. The specific binding of rhIL-1beta to the apical surface of MDCK cells, which was defined as the difference between the amounts of radioactivity bound in the absence and presence of excess unlabeled rhIL-1beta , had reached a steady state after incubation for 4 h at 4 °C (Fig. 4, bullet ). In contrast, the extent of specific binding to the basolateral surface was low throughout the incubation period for up to 3 h (Fig. 4B). The specific binding was not observed in assays of non-polarized MDCK cells cultured in monolayers (data not shown), suggesting that rhIL-1beta binds to GPI-anchored proteins on the apical membranes.


Fig. 4. Time course of 35S-rhIL-1beta binding to apical (A) and basolateral (B) surfaces of MDCK cells at 4 °C. The polarized cell monolayers were prepared as described under "Experimental Procedures." Either the apical or basolateral side of cell monolayers was treated with 50 µl of cold MEM containing approximately 0.2 µCi/ml 35S-rhIL-1beta with or without excess unlabeled rhIL-1beta , and the cells were incubated for the indicated times at 4 °C. The cell monolayers were then washed and harvested as described under "Experimental Procedures." black-square, total binding; bullet , specific binding; open circle , nonspecific binding.
[View Larger Version of this Image (22K GIF file)]


Scatchard Plot Analysis of 35S-rhIL-1beta Binding to MDCK Cells

Fig. 5 shows Scatchard plots (29) for 35S-rhIL-1beta binding to the apical cell surface of polarized MDCK monolayers, where B is moles of bound rhIL-1beta /dpm as a proportion of the total rhIL-1beta /well and F is moles of unbound rhIL-1beta . The plots gave a straight line, indicating homogeneity with respect to affinity for the GPI-anchored glycan. The Ka value for binding was calculated to be 4.6 × 107 M-1. These values were several orders lower than those determined for both type I and type II IL-1 receptors (30). Although both types of IL-1 receptors are expressed in several cell lines, including B lymphoblastoid lines (30), HepG2 (30), and thymoma lines (31), the expression of IL-1 receptors in renal epithelial cell lines has not been reported. The specific binding shown in Fig. 4 indicated that rhIL-1beta binds to neither type I nor type II IL-1 receptors.


Fig. 5. 35S-IL-1beta -specific binding to the apical surface of MDCK cells. The experimental conditions were the same as those in the legend to Fig. 4, except that the apical surface of cell monolayers was treated with 50 µl of MEM containing the indicated concentrations of 35S-IL-1beta with or without excess unlabeled IL-1beta , and the cells were incubated for 4 h at 4 °C. A, the amount of specifically bound 35S-IL-1beta was determined by subtracting the radioactivity bound in the presence of unlabeled IL-1beta from the radioactivity bound in the absence of unlabeled IL-1beta ; B, 35S-IL-1beta binding to the apical surface was plotted by the method of Scatchard (29).
[View Larger Version of this Image (16K GIF file)]


Inhibitory Effect of Mannose 6-Phosphate, Mannose 1-Phosphate, or Mannose 6-Sulfate on 35S-rhIL-1beta Binding to MDCK Cells

To determine whether the ligands required for rhIL-1beta binding to polarized MDCK cells are present within the glycan portion of GPI-anchored proteins, the effect of various amounts of mannose 6-phosphate, mannose 1-phosphate, or mannose 6-sulfate on the specific binding was examined (Fig. 6). The specific binding was dose dependently inhibited by the addition of mannose 6-phosphate (Fig. 6, bullet ), which is the common core structure of the glycan portion; in contrast, mannose 1-phosphate (open circle ) or mannose 6-sulfate (black-square) did not show any inhibitory effect up to 1 mM. These results support the view that the mannose 6-phosphate diester component of the glycan portions of GPI-anchored proteins in the polarized MDCK cells is the moiety recognized by rhIL-1beta in the lectin-like binding interaction.


Fig. 6. Effects of mannose 6-phosphate, mannose 1-phosphate, or mannose 6-sulfate on 35S-IL-1beta -specific binding to the apical surface of MDCK cells. The apical surface of MDCK cells in monolayers was assayed for 35S-IL-1beta binding in the presence of the indicated concentrations of the components for 4 h at 4 °C. Other experimental conditions were the same as those in the legend to Fig. 4. bullet , mannose 6-phosphate; open circle , mannose 1-phosphate; black-square, mannose 6-sulfate.
[View Larger Version of this Image (12K GIF file)]



DISCUSSION

This paper demonstrates that rhIL-1beta recognizes mannose 6-phosphate diester within the glycans of GPI-anchored proteins immobilized on plates or immobilized on CNBr-activated Sepharose 4B. It is also shown that 35S-rhIL-1beta binds to intact GPI-anchored proteins present on the apical membranes of polarized MDCK cells with the Ka value of 4.6 × 107 M-1, and this binding is inhibited in the presence of approximately 1 × 10-6 M mannose 6-phosphate.

The lectin-like interaction between rhIL-1beta and mannose 6-phosphate (diester) in the GPI anchor was calcium-independent as shown in Fig. 2C. It is well known that one of two receptors for lysosomal enzymes trafficking into the lysosome is a Ca2+-independent mannose 6-phosphate binding lectin with a molecular weight of 275,000 (32). It recognizes mannose 6-phosphate residues in high mannose-type sugar chains with a Ka value of 7-8 × 106 M-1 (32), but not mannose 6-phosphate diester (33), while IL-1beta recognizes both mannose 6-phosphate and its diester equally as described in this study. Since no homology was observed between the amino acid sequences of the Ca2+-independent mannose 6-phosphate-binding lectin and IL-1beta using the FASTA method (34), IL-1beta may have a different type of carbohydrate binding domain from that of the Ca2+-independent mannose 6-phosphate-binding lectin for lysosomal enzymes.

It has been reported that not only IL-1beta but also TNF-alpha and lymphotoxin bind to uromodulin immobilized on plates, and in each case this binding is partially inhibited by high concentrations of N,N'-diacetylchitobiose (7, 8). Since uromodulin has not only asparagine-N-linked sugar chains but also a GPI-anchored glycan at the carboxyl-terminal end (35), it seems important to ascertain in the near future whether TNF-alpha and lymphotoxin have the same carbohydrate binding specificities as IL-1beta .

Although the physiological function of the lectin-like character of these cytokines remains to be studied, several possible roles may be considered as follows. It has been reported that TNF-alpha exerts a direct trypanolytic activity on salivarian trypanosomes (36). This activity is not blocked by soluble 55-kDa and 75-kDa TNF receptors, but is potently inhibited by N,N'-diacetylchitobiose and the hexapeptide of TNF-alpha , Thr105 to Glu110, implying the existence of an alternative recognition domain of TNF-alpha for constituents of trypanosome parasites (36). It is of interest to determine whether this recognition region of TNF-alpha shows binding activity specific for mannose 6-phosphate diester in the glycan of the GPI anchor, i.e. the same lectin-like specificity as IL-1beta . Such a lectin-like activity of TNF-alpha may exhibit such alternative recognition of microbial constituents.

As another physiological role of the lectin-like interaction between rhIL-1beta and the GPI anchor, the latter may function as a modulator of IL-1 receptors. Although two types of IL-1 receptors have been identified (37), it remains to be established whether the cloned IL-1 receptor by itself has the ability to generate signals or whether it is linked to an unidentified signal transducer or effector. It is clear only that the two different IL-1 receptors function independently at the level of ligand binding and do not form a heterodimeric receptor complex even when they are present on the same cell (37). The GPI anchor portions that are recognized with low affinity by IL-1beta may serve only to bind ligand and deliver it or enhance binding to the type I and/or II receptor(s), which would be capable of binding ligands and delivering the signal into cells, as in the case of the interleukin-2 system (38) or growth factors such as granulocyte-macrophage colony-stimulating factor (2), heparin-binding growth factor (3), basic fibroblast growth factor (4, 5), or the midkine-heparan sulfate interaction (6). The growth factors show low affinity for heparan sulfate, which behaves as an accessory molecule required for binding of growth factors to the high affinity receptors on the plasma membrane. Such an obligatory relationship of low and high affinity cytokine receptors suggests a physiological role for heparan sulfate and the GPI anchor as low affinity receptors and constitutes a novel mechanism for the regulation of cytokine-mediated signal transduction.

It is reported that some IL-1beta -specific binding structures with low affinity are present on the cell wall of virulent bacteria (39) and in supernatants from vaccinia virus-infected CV-1 cells (40). In these cases, the role of binding to IL-1beta is thought to prohibit or diminish the inflammatory effects of the cytokine produced by infected cells. The low affinity binding of the GPI anchor portion to IL-1beta may have a similar role in regulating the effects of IL-1beta , which are induced by binding of IL-1beta to type I and/or II receptor(s).

The results so far described support the view that these lectin-like activities of cytokines are fundamental to immunology and bring some additional information to light concerning the functional role of the GPI anchor.


FOOTNOTES

*   This work was supported by Grant-in-Aid 06240246 for Scientific Research on Priority Areas from the Ministry of Education, Science, and Culture, Japan.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.: 81-3-3294-3286; Fax: 81-3-3294-2656.
1   The abbreviations used are: TNF, tumor necrosis factor; IL-1beta , interleukin-1beta ; rhIL-1beta , recombinant human interleukin-1beta ; IL-1, interleukin-1; VSG, variant surface glycoproteins; PI-PLC, phosphatidylinositol-specific phospholipase C; GPI, glycosylphosphatidylinositol; ALP, human placental alkaline phosphatase; BSA, bovine serum albumin; PBS, phosphate-buffered saline; MDCK, Madin-Darby canine kidney cells; MEM, minimal essential medium.

REFERENCES

  1. Arai, K., Lee, F., Miyajima, A., Miyatake, S., Arai, N., and Yokota, T. (1990) Annu. Rev. Biochem. 59, 783-836 [CrossRef][Medline] [Order article via Infotrieve]
  2. Gordon, M. Y., Riley, G. P., Watt, S. M., and Greaves, M. F. (1987) Nature 326, 403-405 [CrossRef][Medline] [Order article via Infotrieve]
  3. Rosengart, T. K., Johnson, W. V., Friesel, R., Clark, R., and Maciag, T. (1988) Biochem. Biophys. Res. Commun. 152, 432-440 [Medline] [Order article via Infotrieve]
  4. Bashkin, P., Doctrow, S., Klagsbrun, M., Svahn, C., Folkman, J., and Vlodavsky, I. (1989) Biochemistry 28, 1737-1743 [Medline] [Order article via Infotrieve]
  5. Seno, M., Sasada, R., Kurokawa, T., and Igarashi, K. (1990) Eur. J. Biochem. 188, 239-245 [Abstract]
  6. Tomomura, M., Kadomatsu, K., Nakamoto, M., Muramatsu, H., Kondoh, H., Imagawa, K., and Muramatsu, T. (1990) Biochem. Biophys. Res. Commun. 171, 603-609 [Medline] [Order article via Infotrieve]
  7. Muchmore, A. V., and Decker, J. M. (1987) J. Immunol. 138, 2541-2546 [Abstract/Free Full Text]
  8. Fukushima, K., Watanabe, H., Takeo, K., Nomura, M., Asahi, T., and Yamashita, K. (1993) Arch. Biochem. Biophys. 304, 144-153 [CrossRef][Medline] [Order article via Infotrieve]
  9. Muchmore, A. V., Shifrin, S., and Decker, J. M. (1987) J. Immunol. 138, 2547-2553 [Abstract/Free Full Text]
  10. Sherblom, A. P., Sathyamoorthy, N., Decker, J. M., and Muchmore, A. V. (1989) J. Immunol. 143, 939-944 [Abstract/Free Full Text]
  11. Vehmeyer, K., Brandt, W., Nagel, G. A., and Gabius, H. J. (1988) Life Sci. 43, 1591-1598 [Medline] [Order article via Infotrieve]
  12. Yamashita, K., Ichishima, E., Arai, M., and Kobata, A. (1980) Biochem. Biophys. Res. Commun. 96, 1335-1342 [Medline] [Order article via Infotrieve]
  13. Yamashita, K., Ohkura, T., Okada, S., Yabuuchi, H., and Kobata, A. (1981) J. Biol. Chem. 256, 4789-4798 [Abstract]
  14. Ogata, S., Hayashi, Y., Takami, N., and Ikehara, Y. (1988) J. Biol. Chem. 263, 10489-10494 [Abstract/Free Full Text]
  15. Pannecoucke, X., Schmitt, G., and Luu, B. (1994) Tetrahedron 50, 6569-6578 [CrossRef]
  16. Marugg, J. E., Tromp, M., Yeheskiely-Kuyl, E., van der Marel, G. A., and van Boom, J. H. (1986) Tetrahedron Lett. 27, 2661-2664 [CrossRef]
  17. Yamanoi, T., Nakamura, K., Takeyama, H., Yanagihara, K., and Inazu, T. (1994) Bull. Chem. Soc. Jpn. 67, 1359-1366
  18. Tamm, I., and Horsfall, F. L., Jr. (1952) J. Exp. Med. 95, 71-97
  19. Cross, G. A. (1975) Parasitology 71, 393-417 [Medline] [Order article via Infotrieve]
  20. Laemmli, U. K. (1970) Nature 227, 680-685 [Medline] [Order article via Infotrieve]
  21. Joseph-Liauzun, E., Leplatosis, P., Legoux, R., Guerveno, V., Marchese, E., and Ferrara, P. (1990) Gene (Amst.) 86, 291-295 [Medline] [Order article via Infotrieve]
  22. Baclarova-Stander, J., Pfeiffer, S. E., Fuller, S. D., and Simon, K. (1984) EMBO J. 3, 2687-2694 [Abstract]
  23. Spik, G., Bayard, B., Fournet, B., Strecker, G., Bouquelet, S., and Montreuil, J. (1975) FEBS Lett. 50, 296-299 [CrossRef][Medline] [Order article via Infotrieve]
  24. Liang, C-J., Yamashita, K., and Kobata, A. (1980) J. Biochem. (Tokyo) 88, 51-58 [Abstract]
  25. Redman, C. A., Thomas-Oates, J. E., Ogata, S., Ikehara, Y., and Ferguson, M. A. (1994) Biochem. J. 302, 861-865 [Medline] [Order article via Infotrieve]
  26. Ferguson, M. A., Low, M. G., and Cross, G. A. (1985) J. Biol. Chem. 260, 14547-14555 [Abstract/Free Full Text]
  27. Endo, T., Ohbayashi, H., Hayashi, Y., Ikehara, Y., Kochibe, N., and Kobata, A. (1988) J. Biochem. (Tokyo) 103, 182-187 [Abstract]
  28. Lisanti, M. P., Sargiacomo, M., Graeve, L., Saltiel, A. R., and Rodriguez, B. E. (1988) Proc. Natl. Acad. Sci. U. S. A. 85, 9557-9561 [Abstract]
  29. Scatchard, G. (1949) Ann. N. Y. Acad. Sci. 51, 600-672
  30. McMahan, C. J., Slack, J. L., Mosley, B., Cosman, D., Lupton, S. D., Brunton, L. L., Grubin, C. E., Wignall, J. M., Jenkins, N. A., Brannan, C. I., Copeland, N. G., Huebner, K., Croce, C. M., Cannizzarro, L. A., Benjamin, D., Dower, S. K., Spriggs, M. K., and Sims, J. E. (1991) EMBO J. 10, 2821-2832 [Abstract]
  31. Sims, J. E., March, C. J., Cosman, D., Widmer, M. B., MacDonald, H. R., McMahan, C. J., Grubin, C. E., Wignall, J. M., Jacson, J. L., Call, S. M., Friend, D., Alpert, A. R., Gillis, S., Urdal, D. L., and Dower, S. K. (1988) Science 241, 585-589 [Medline] [Order article via Infotrieve]
  32. Kornfeld, S. (1992) Annu. Rev. Biochem. 61, 307-330 [CrossRef][Medline] [Order article via Infotrieve]
  33. Li, M. M., Distler, M. L., and Jourdian, G. W. (1989) Methods Enzymol. 179, 304-310 [Medline] [Order article via Infotrieve]
  34. Pearson, W. R., and Lipman, D. J. (1988) Proc. Natl. Acad. Sci. U. S. A. 85, 2444-2448 [Abstract]
  35. Rindler, M. J., Naik, S. S., Li, N., Hoops, T. C., and Peraldi, M. N. (1990) J. Biol. Chem. 265, 20784-20789 [Abstract/Free Full Text]
  36. Lucas, R., Magez, S., De Leys, R., Fransen, L., Scheerlinck, J-P., Rampelberg, M., Sablon, E., and De Baetselier, P. (1994) Science 263, 814-817 [Medline] [Order article via Infotrieve]
  37. Slack, J., McMahan, C. J., Waugh, S., Schooley, K., Spriggs, M. K., Sims, J. E., and Dower, S. K. (1993) J. Biol. Chem. 268, 2513-2524 [Abstract/Free Full Text]
  38. Hatakeyama, M., Tsudo, M., Minamoto, S., Kono, T., Doi, T., Miyata, T., Miyasaka, M., and Taniguchi, T. (1989) Science 244, 551-556 [Medline] [Order article via Infotrieve]
  39. Porat, R., Clark, B. D., Wolff, S. M., and Dinarello, C. A. (1991) Science 254, 430-432 [Medline] [Order article via Infotrieve]
  40. Spriggs, M. K., Hruby, D. E., Maliszewski, C. R., Pickup, D. J., Sims, J. E., Buller, R. M., and Vanslyke, J. (1992) Cell 71, 145-152 [Medline] [Order article via Infotrieve]

©1997 by The American Society for Biochemistry and Molecular Biology, Inc.