Advertisement
JBC

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Originally published In Press as doi:10.1074/jbc.M205282200 on July 1, 2002

J. Biol. Chem., Vol. 277, Issue 41, 38764-38771, October 11, 2002
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
277/41/38764    most recent
M205282200v1
Right arrow Submit a Letter to Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Zanetta, J.-P.
Right arrow Articles by Vergoten, G.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Zanetta, J.-P.
Right arrow Articles by Vergoten, G.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

Evidence for a Lectin Activity for Human Interleukin 3 and Modeling of Its Carbohydrate Recognition Domain*

Jean-Pierre ZanettaDagger §, Roland BindeusDagger , Guy Normand, Viviane DurierDagger , Philippe LagantDagger , Emmanuel MaesDagger , and Gérard VergotenDagger

From the Dagger  CNRS Unité Mixte de Recherche 8576, Glycobiologie Structurale et Fonctionnelle, Université des Sciences et Technologies de Lille Bâtiment C9, 59655 Villeneuve d'Ascq Cedex, France and the  Physiologie Cellulaire et Moléculaire de la Rétine, INSERM EPI9918, Hôpitaux Universitaire de Strasbourg, 1 place de l'Hôpital, 67091 Strasbourg Cedex, France

Received for publication, May 29, 2002, and in revised form, June 27, 2002

    ABSTRACT
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES

We demonstrate that human interleukin 3 (IL-3) is a lectin recognizing specifically the glycosaminoglycan part of a chondroitin sulfate proteoglycan (PGS3; Normand, G., Kuchler, S., Meyer, A., Vincendon, G., and Zanetta, J. P. (1988) J. Neurochem. 51, 665-676) isolated from the adult rat brain. The specificity of the interaction of this particular proteoglycan with IL-3 is due to the abundance of GlcA(2S)beta 1,3GalNAc(4S)beta 1 disaccharide units as suggested by 1H NMR. Computational docking experiments of the lower energy conformers of the different disaccharides from chondroitin sulfates reveal a privileged binding site for GlcA(2S)beta 1,3GalNAc(4S)beta 1 (involving His-26, Arg-29, Asn-70, and Trp-104) localized in an area of IL-3 different from the receptor-binding domain previously identified by others (Bagley, C. J., Phillips, J., Cambareri, B., Vadas, M. A., and Lopez, A. F. (1996) J. Biol. Chem. 271, 31922-31928). Molecular modeling of the mutation P33G, described as increasing the biological activity of IL-3 without affecting its receptor binding (Lokker, N. A., Movva, N. R., Strittmatter, U., Fagg, B., and Zenke, G. (1991) J. Biol. Chem. 266, 10624-10631) provokes a change of the three-dimensional structure of IL-3, especially in the area of the putative carbohydrate recognition domain defined above. Computational docking experiments of the different disaccharides of chondroitin sulfates indicate a loss of affinity for the previous ligand but a higher affinity for the classic disaccharide of chondroitin-4-sulfate. This change from a rare and specific ligand to a more abundant constituent of proteoglycans could induce an increased quantitative association between the IL-3 receptors and its ligands and, consequently, an increased signaling.

    INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES

Interleukin 3 (IL-3) is a member of the cytokine superfamily that promotes multipotential hematopoietic cell growth by interacting with a cell surface receptor composed of alpha - and beta -chains and possessing a domain necessary for signaling but not for receptor binding. We made the hypothesis that this domain could be a carbohydrate recognition domain (CRD)1 as evidenced for several interleukins (1, 2) able to specifically associate the interleukin receptor with other surface complexes expressing the carbohydrate ligand of the interleukin. In previous experiments (1),2 none of the tested cytokines was able to bind significantly to mixtures of commercially available glycosaminoglycans (GAG). These negative results were surprising because several studies (3-5) reported interactions of cytokines with proteoglycans. The negative results of our previous studies could, at least in part, be explained by the use of a different methodology. In fact, our method (1) measured by immunoblotting the quantity of unlabelled recombinant human cytokine unbound to plastic microwells containing or not containing immobilized putative ligands. Consequently, binding could be observed only when the ligand was present at a stoichiometric level with the cytokine and when the affinity of the cytokine for its ligand was sufficiently high.

Because GAGs present large structural varieties and because the commercial GAGs used in the previous study (1) were isolated from non-mammalian organisms, we decided to test the binding of different cytokines to different immobilized proteoglycans previously isolated from the adult rat brain (6).

This study demonstrates the specific binding of IL-3 to a chondroitin sulfate proteoglycan isolated from adult rat brain. This compound possesses an unusually high degree of sulfation, with the majority of the disaccharide units formed of 2-O-sulfated glucuronic acid and of 4-O-sulfated N-acetylgalactosamine as suggested by 1H NMR of its glycosaminoglycan part. Computational docking of the lower energy conformers of the disaccharides from chondroitin-4-sulfate and chondroitin-6-sulfate, chondroitin-4,6-disulfate, chondroitin-2,4-disulfate, and chondroitin-2,6-disulfate results in a specific docking of GlcA(2S)beta 1,3GalNAc(4S)beta 1 in the area of Trp-104 previously proposed as a domain of IL-3 necessary for its function. Molecular modeling of the mutation P33G that gives a mutein with increased biological activity (7, 8) indicates a complete modification of this domain (with the complete accessibility of Trp-104). In silico docking of all disaccharides described above indicates that the theoretical best ligand is no more GlcA(2S)beta 1,3GalNAc(4S)beta 1 but the classic disaccharide derived from chondroitin-4-sulfate. These data suggest that the increased activity of the P33G mutein is due to a change from a rare disaccharide ligand to a common disaccharide ligand, allowing the increase of IL-3-dependent signaling.

    MATERIALS AND METHODS
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES

Chemicals-- Interleukin 1alpha (IL-1alpha ), -1beta , -3, -5, -7 and -8, interferon-gamma (IFNgamma ), and their respective polyclonal antibodies were purchased from Chemicon (Temucula, CA). Monoclonal anti-rabbit IgG, chondroitin sulfate A, B, and C, heparan sulfate, heparin, hyaluronic acid, keratan sulfate, bovine serum albumin, nitro blue tetrazolium (NBT), 5-bromo-4-chloro-3-indolyl-phosphate (BCIP), and cetyl-pyridinium chloride (CPC) were from Sigma. Nitrocellulose (0.45 µm pore size) was from Schleicher & Schuell (Darmstadt, Germany). D2O (99.97% purity; Euriso-top®) was from CEA (Saclay, France). Periodate-treated bovine serum albumin was prepared according to Glass et al. (9).

Isolation of Rat Brain Proteoglycans-- In typical preparations, 240 forebrains (from adult Wistar albino rats) were homogenized (5% w/v) in 10 mM Tris-HCl buffer, pH 7.2, containing protease inhibitors (10 mM EDTA, 100 mM 6-amino-hexanoate, and 5 mM benzamidine hydrochloride) and centrifuged for 2 h at 100,000 × g. The supernatant and the particulate fractions were treated separately.

The particulate fraction (M fraction) was delipidized (10) and extracted by stirring for 24 h at 4 °C in 0.1 M sodium citrate buffer, pH 3.1, containing the protease inhibitors as above. After centrifugation for 1 h at 23,000 × g, the supernatant was discarded, and the pellet was extracted for 48 h at 4 °C in 50 mM Tris-HCl buffer, pH 7.5, containing 4 M guanidinium chloride and protease inhibitors as above. After centrifugation (1 h at 15,000 × g), the pellet was discarded, and the supernatant was dialyzed against large volumes of cold water. The precipitate was recovered by centrifugation (1 h at 23,000 × g) and extracted again in 0.4 M sodium citrate buffer, pH 5.0, over 24 h at 4 °C. After centrifugation, the supernatants were passed on a DE-52 column (17 × 2.5 cm) previously equilibrated in 0.1 M sodium citrate buffer, pH 5.0. Proteoglycans were eluted with a linear gradient of NaCl from 0 to 1.0 M with a total volume of 200 ml at a flow rate of 30 ml/min. The fractions containing proteoglycans were dialyzed against cold water and lyophilized. After solubilization in a small volume of water, the proteoglycans were precipitated with ethanol (11). The pellet was homogenized in 4 ml of 10 mM Tris-HCl buffer containing 0.4 M NaCl and centrifuged (30 min at 10,000 × g) at room temperature. The supernatant was submitted to gel filtration.

The supernatant fraction of the first extract (S fraction) was precipitated with 0.1% CPC for 24 h at 4 °C (12), and the supernatant was discarded. The pellet was extracted in 50 mM Tris-HCl, pH 7.5, containing 4 M guanidinium chloride and protease inhibitors. Subsequent procedures were identical to those for the M fraction.

The material enriched in proteoglycans was fractionated by gel filtration on an Ultrogel AcA22 column (100 × 2.5 cm) previously equilibrated in 10 mM Tris-HCl buffer, pH 7.5, containing 0.4 M NaCl. The elution was followed by measuring the absorbance at 280 nm and using the carbazole technique (13). Seven different proteoglycans were isolated, four from the M fraction (PGM1, PGM2, PGM3, and PGM4) and three from the S fraction (PGS1, PGS2, and PGS3). The nature of these compounds was tested using enzymatic degradation (chondroitinase AC and ABC), nitrous acid treatment (6), and gas chromatography/mass spectrometry (GC/MS) of heptafluorobutyrate derivatives of the mono- and disaccharides liberated after acid-catalyzed methanolysis in the presence of barium acetate (14).

For preliminary experiments of binding of cytokines, another preparation of proteoglycans was used, consisting of the material obtained applying the previous procedure without initial separation of the M and S fractions and without separation by gel filtration.

Immobilization of Proteoglycans on Plastic Microwells-- Mixtures of proteoglycans or individual compounds were dissolved or suspended in excess (1 mg/ml) in PBS (25 mM sodium phosphate buffer, pH 7.2, containing 150 mM NaCl) and 50 µl were added to each microwell. The liquid was evaporated at 50 °C in an oven. The wells were rinsed three times with PBS and then saturated three times for 2 h at room temperature with 300 µl of a 3% solution in PBS of bovine serum albumin previously treated with periodate (9). This treatment eliminated interferences due to the presence of glycoconjugates in commercially available bovine serum albumin preparations. The wells were rinsed three times with PBS and then water. Before use, the wells were saturated again as above and rinsed three times with PBS. Cytokines were added (100 ng in 50 µl of PBS containing 0.3% periodate-treated bovine serum albumin) and left for 2 h at room temperature under rotary agitation. The supernatants were recovered, supplemented with 25 µl of Laemmli (15) dissociating buffer, boiled for 10 min, and submitted to 13% acrylamide SDS-PAGE in the Laemmli buffer system. The gels were blotted onto nitrocellulose (16). The cytokines were revealed using their respective polyclonal anti-cytokine antibodies, followed by alkaline phosphatase (AKP)-labeled anti-rabbit IgG, and finally NBT/BCIP detection of the AKP activity (17).

Purification of the PGS3-derived Glycosaminoglycan-- Purified PGS3 was submitted to mild alkaline hydrolysis (0.1 M NaOH overnight at 4 °C with stirring) in order to liberate O-linked glycans (GAGs and eventually O-glycans). The sample was neutralized with 0.5 M HCl and then dialyzed (24 h at 4 °C) against precooled distilled water. The dialysate and the remaining samples were dried under vacuum at room temperature in a rotary evaporator. The material retained in the dialysis tube (likely containing GAGs and proteins) was centrifuged, and the supernatant was precipitated with 0.1% CPC overnight at 4 °C. After centrifugation, the pellet was suspended in water, and excess CPC was extracted three times with isoamyl alcohol. The remaining aqueous phase was evaporated to dryness, and the sample was exchanged against D2O for 1H NMR analysis.

NMR Analysis of the PGS3 Glycosaminoglycan-- Prior to NMR spectroscopic analysis, the sample was repeatedly exchanged in D2O with intermediate freeze-drying and then dissolved in 250 µl of D2O. The sample was analyzed in 200 × 5 mM BMS-005-B Shigemi® tubes at 300 and 315 K on a Bruker ASX400-NB spectrometer (1H 400.33MHz, Centre Commun de Mesure RMN, Villeneuve d'Ascq, France) equipped with a double resonance (1H/X) broadband inverse z-gradient probe head. Data were recorded without sample spinning, and chemical shifts were expressed in ppm downfield from the acetone proton signal (Delta 1H, 2.225 ppm).

Molecular Modeling of the Disaccharide Ligands of IL-3-- The structures of low energy conformers of the disaccharide ligand of IL-3 were obtained using the minimization and random searching tools within the SYBYL 2002 version (SYBYL, Tripos Inc., St Louis, MO; www.tripos.com). Parameters for the internal rotation potentials are those obtained in our group.3 The electrostatic charges were derived from quantum mechanical calculations using the Jaguar software (Jaguar version 4.0, Schrödinger Inc., Portland, OR; www.schrodinger.com). The density functional theory was used at the B3LYP/6-31G* level. For that purpose, the molecular electrostatic potential has been fitted to a set of point charges located at the atomic centers, reproducing the dipole moment. The electrostatic potential itself was computed on a spherical grid. A random conformational search technique was used to sample the conformational space. The two glycosidic linkage rotatable bonds and five rotatable bonds for substituents were defined. The structures of low energy conformers were then compared with those of disaccharides of classic chondroitin sulfate. In all cases, 20,000,000 configurations were examined using bump checking (with a van der Waals scaling factor of 0.7) and energy as criteria.

In Silico Docking of the Chondroitin Sulfate Disaccharides in the IL-3 Molecule-- For the five lower energy conformers of the five disaccharides of chondroitin sulfates, a surface docking with IL-3 (PDB 1JL1) has been investigated. The Global Range Molecular Matching (GRAMM, 2002 version) program was used (18). This technique locates the area of the global minimum of intermolecular energy for the structure, with different accuracy. The generic mode and a grid step of 1.7 were used. 1000 matches were stored, among which 100, 10, and 5 were displayed. For the lower energy disaccharide/IL-3 complexes, a dynamic energy-based optimization was performed. The data of the conformations of the disaccharides and of the docking experiments were analyzed as PDB files using the Swiss-Pdb-Viewer (www.expasy.ch) and/or the Weblab ViewerLite 4.0 and 4.2 (www.msi.com). For the computational study of the possible modification of the three-dimensional structure of IL-3 under the effect of specific mutations, amino acids were altered using the mutate option of the SYBYL software, followed by the classic program of energy minimization for the whole molecule.

    RESULTS
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES

Using the technique described above (see "Materials and Methods"), none of the cytokines bound to a mixture of commercially available glycosaminoglycans (not shown; see Ref. 1 for IL-1alpha , -1beta , -4, -6, and -7). Because binding of cytokines to proteoglycans was reported, we tested the binding of these cytokines to a mixture of proteoglycans isolated from the adult rat brain (6). As shown in Fig. 1A, the different cytokines showed different behavior. IL-1beta , -5, and -7 did not show any binding. The same was found for IL-1alpha and interferon-gamma despite the low efficiency of the antibodies used here. The behavior of IL-8 could not be studied; the antibody always gave negative results even when IL-8 was directly submitted to SDS-PAGE and blotting, probably because of the specific antibody reactivity against the non-denatured form of IL-8. In contrast, IL-3 showed an important binding to this mixture, which was considered as specific because it was not observed in wells lacking the proteoglycans (Fig. 1A) and was the only cytokine showing a strong binding.


View larger version (66K):
[in this window]
[in a new window]
  		Fig. 1.   Immunoblotting study of the fixation of different cytokines to the mixture of rat brain proteoglycans (A), the fixation of IL-3 to the different isolated rat brain proteoglycans (B), and inhibition of the fixation to PGS3 (C). In A, B, and C, + corresponds to wells containing immobilized proteoglycans and - to wells not containing proteoglycans. A, note the complete fixation of 100 ng IL-3 to the proteoglycan mixture. B, note the complete fixation of IL-3, to 1 µg of immobilized PGS3, and the total absence of IL-3, to 10 µg of immobilized PGM1, PGM3, PGS1, and PGS2. IL-3 showed a partial binding to 10 µg of immobilized PGM2. C, binding to PGS3 (lane T, +) was not inhibited by the dialyzable oligosaccharides (lane 1, +) but with 10 µg of the purified GAG portion of PGS3 (lane 2, +). The quantity of IL-3 was equivalent to that recovered in wells uncoated with PGS3 (lane T, -).

The different behavior of IL-3 on mixtures of GAGs and on mixtures of different proteoglycans was surprising, because the proteoglycan mixture actually contained chondroitin sulfate A and B, keratan sulfate, heparan sulfate, and hyaluronic acid (the latter found as a contaminant in the preparations). Because this mixture contained seven different proteoglycans (6), four derived from the membrane fraction (PGM1, PGM2, PGM3, and PGM4) and three from the soluble fraction (PGS1, PGS2, and PGS3) of adult rat brain, these components were individually immobilized on plastic microwells. When the wells were supplemented with 100 ng of IL-3, 20 and 90% of IL-3 were bound to PGM2 and PGS3, respectively (Fig. 1B). The binding to PGM1, PGM3, PGM4, PGS1, and PGS2 was negligible in contrast with PGS3 (90 ng for 1 µg of immobilized PGS3, instead of 20 ng for 10 µg of immobilized PGM2).

Enzymatic and chemical degradations (6) and GC/MS analysis (14) allowed precise measurement of the specificity of the binding of IL-3. Indeed, PGM1 was a mixture of hyaluronic acid and a proteoglycan containing heparan sulfate and a chondroitin sulfate chain, whereas PGM2, PGM3, and PGM4 were similar but were not contaminated with hyaluronic acid. In contrast, PGS1, PGS2, and PGS3 were all chondroitin sulfate proteoglycans devoid of heparan sulfates and dermatan sulfate, as verified by the complete degradation of the GAG portion using chondroitinase AC (6). Therefore, it was suggested that the ligand of IL-3 was a chondroitin sulfate proteoglycan. The absence of characteristic disaccharides released from heparan sulfates and the absence of iduronic acid, together with the unique presence of disaccharides derived from chondroitin sulfates, observed by GC/MS (14) verified these conclusions.

Based on chemical analysis (6), PGS3 differed from the other proteoglycans by its higher ratio of sulfate relative to GalNAc (1.63/1.00), between 0.34 for PGM3 and 0.81 for PGS2. Throughout the different proteoglycans of the M fraction, PGM2 also showed a higher degree of sulfation. This higher level of sulfation might explain the specific binding of IL-3 to these particular chondroitin sulfate brain proteoglycans and not to the others, as well as to commercially available A and C chondroitin sulfates.

Inhibition of IL-3 Binding to PGS3 by the GAG Portion of PGS3-- To determine the nature of the interaction between PGS3 and IL-3, we performed inhibition experiments of the binding with glycans isolated from PGS3. In fact, the monosaccharide analysis of PGS3 (6) revealed the presence of a significant amount of mannose in the molecule, indicating the presence of O- or N-linked glycans. Consequently, it was necessary to confirm that the binding of IL-3 to PGS3 was not because of these glycans. Therefore, after a mild beta -elimination procedure, the small oligosaccharides were dialyzed and separated from the non-dialyzable material, including chondroitin sulfate chains and the protein portion (possibly associated with N-glycans). The two different fractions were tested for their inhibitory properties for the binding of IL-3 to the immobilized intact PGS3. As shown in Fig. 1C, the dialyzed material was not inhibitory, in contrast with the non-dialyzable material (not shown). The inhibitory material was purified by precipitation with CPC. As shown in Fig. 1C, only the GAG portion of PGS3 was inhibitory. Therefore, these experiments demonstrated that the high affinity ligand of IL-3 was the GAG portion of PGS3. The specificity of the binding of IL-3 to PGS3 was likely due to its peculiar sulfation pattern, because chondroitin 4 and 6 sulfates from whale cartilage (Sigma) were inactive.

Determination of the Position of the Sulfate Groups of PGS3-- To determine the positions of the sulfate groups in the chondroitin chain, the glycosaminoglycan part of PGS3 liberated by mild alkaline hydrolysis (see "Materials and Methods") was submitted to 1H NMR analysis at 300 and 315 K. The chemical shifts of different protons indicated that, despite an heterogeneity, the majority of the material was composed of GlcA sulfated in position 2 and of GalNAc sulfated in position 4. Indeed, no signal was detected for the H2 protons of beta GlcA at 3.375-3.379 ppm (19-22) characteristic of non-sulfated GlcA and around 3.64 ppm characteristic of 3-sulfated GlcA. In contrast, the characteristic downfield-shifted signal at 4.11 ppm of 2-sulfated GlcA (20) was observed. This conclusion was reinforced by the presence of the downfield-shifted signal (4.80 instead of 4.45 ppm) of the anomeric proton of GlcA. This indicated that at least 90% of the GlcA residues were sulfated in position 2. For GalNAc, two signals corresponding to the H4 protons were detected at 4.19 and 4.75 ppm in the proportion 1:2, indicating the presence of both non-sulfated GalNAc at 4.19 ppm and 4-sulfated GalNAc at 4.75 ppm (22). Therefore, these data indicated that the major disaccharide repetitive units of PGS3 were GlcA(2S)beta 1,3GalNAc(4S)beta 1-, the other being GlcA(2S)beta 1,3GalNAcbeta 1- (in a molar ratio of 2:1).

Molecular Modeling of the Disaccharide Ligands of IL-3-- The conformations of the lower energy conformers of the disaccharide ligand of IL-3 GlcA(2S)beta 1,3GalNAc(4S)beta 1 were calculated and compared with the conformations of the disaccharides of classic chondroitin sulfate A, GlcAbeta 1,3GalNAc(4S)beta 1, and chondroitin sulfate C, GlcAbeta 1,3GalNAc(6S)beta 1, and with those of GlcA(2S)beta 1,3GalNAc(6S)beta 1 and of GlcAbeta 1,3GalNAc(4,6S)beta 1. As shown in Fig. 2, the five lower energy conformers of the disaccharide ligands of IL-3, GlcA(2S)beta 1,3GalNAc(4S)beta 1, showed two groups of compounds, with the different energy in each group caused by the presence of different intramolecular hydrogen bonds. Compound A in Fig. 2 showed the lowest energy (0.7 kcal·mol-1 lower than compound B and 1.2 kcal·mol-1 lower than the three other conformers). These five lower energy conformers differed from the others by a jump in energy of higher than 5 kcal·mol-1.


View larger version (31K):
[in this window]
[in a new window]
  		Fig. 2.   Calculated conformations of the five lower energy conformers of the GlcA(2S)beta 1,3GalNAc(4S)beta 1 disaccharide. Compound A is the lower energy conformer. Compounds B, C, D, and E showed increased energies of 0.76, 1.233, 1.395, and 2.105 kcal·mol-1 relative to compound A.

In Silico Docking of the Ligands of IL-3 in the IL-3 Molecule-- For the five families of disaccharides, the five lower energy conformers were submitted to computational docking into the three-dimensional structure of IL-3, accessible as PDB 1JL1. The non-ligand oligosaccharides showed a quasi-random distribution with a predominance of localization in the area containing His-95 and -98 (not shown), with fine analysis showing that the major interactions were of ionic nature. Because the calculations were performed in vacuo (epsilon  = 1), these types of interactions were considered very unlikely to take place in vivo. The absence of van der Waals interactions and/or hydrogen bonds with the surrounding amino acids reinforced this conclusion. This was not the case for the lower energy conformers of GlcA(2S)beta 1,3GalNAc(4S)beta 1 because 70% of them were found in the area of His-26, Arg-29, Asn-70, and Trp-104 (Fig. 3A). When the experiments included only the 10 energy conformers of the IL-3 ligand, the partition became 3 to 7 in the same sites (Fig. 3B). This ratio was 1:4 for the five lower energy conformers of the ligand of IL-3 (Fig. 3C).


View larger version (42K):
[in this window]
[in a new window]
  		Fig. 3.   Views of the computational docking of the 100 (A), 10 (B), and 5 (C) lower energy complexes of the lower energy conformer (Fig. 2, compound A) of GlcA(2S)beta 1,3GalNAc(4S)beta 1 into the three-dimensional structure of human IL-3 (PDB 1JL1).

The fine analysis of the results of the docking experiments of the five lower energy conformers indicated a preferential localization of four of them in close apposition to the Trp-104 residue. The exception was one that localized in the area of His-95 and -98 (Fig. 3C). As discussed above, this site was rejected as a putative CRD because of the predominance of ionic bonds and the absence of van der Waals interactions. The other conformers all localized in the same region, and all interacted with Trp-104 through strong van der Waals interactions and hydrogen bonding with surrounding amino acids. As shown in Fig. 4A, for the lower energy conformers (shown in Fig. 2) this site was essentially different from that defined (23) as the receptor-binding site of IL-3 (Fig. 4A, blue). The fine analysis of the docking data of the four lower energy conformers of GlcA(2S)beta 1,3GalNAc(4S)beta 1 indicated the same major features, van der Waals interactions with Trp-104 and His-26 and hydrogen bonds with the same amino acids. After energy minimization of the complexes, the lower energy was found for a disaccharide ligand presenting an intermediate conformation between conformers A and B of Fig. 2 without modification of the conformation of the protein.


View larger version (58K):
[in this window]
[in a new window]
  		Fig. 4.   A, representation of the putative carbohydrate recognition domain of human interleukin 3. Trp-104 is shown in yellow, the disaccharide ligands GlcA(2S)beta 1,3GalNAc(4S)beta 1 in green, and amino acids involved in the formation of hydrogen bonds with the ligands in purple. The amino acids proposed as part of the receptor-binding domain (23) are shown in blue. The Pro-33 residue mutated to Gly in B is in red. B, calculated conformation of the P33G mutein and computational docking of GlcAbeta 1,3GalNAc(4S)beta 1- into the mutein. Note that the area of the putative CRD of wild-type IL-3 is completely modified, whereas modifications were less evident in the receptor-binding domain.

The precise interactions of this compound with the structure of IL-3 (Fig. 5A) showed a very intense van der Waals interaction of the pyranic ring of GalNAc facing the indolic group of Trp-104 (2.86-4.16 Å). Four hydrogen bonds stabilized the binding. There was one hydrogen bond (2.68 Å) between the oxygen of the sulfate group of GlcA and the nitrogen atom of the lateral chain of Asn-70, one hydrogen bond (2.61 Å) between the oxygen atom of the sulfate group of GalNAc and the nitrogen atom of the side chain of Asn-70, one hydrogen bond (2.78 Å) between the oxygen of the polypeptide amide bond of His-26 and the oxygen atom of the carboxylic group of GlcA, and one hydrogen bond (2.94 Å) between the nitrogen atom of the guanidinium group of Arg-29 and the oxygen atom of the carboxylic group of GlcA. A weaker van der Waals interaction was also observed between the pyranic ring of GlcA and the side chain of His-26 (4.79-6.59 Å). This configuration (with two series of van der Waals interactions and four strong hydrogen bonds) fulfilled the common criteria defined for the binding of disaccharides in the CRD of calcium-independent lectins (24-26). Furthermore, this site was clearly different from the receptor-binding domain as defined by site-directed mutagenesis experiments. Finally, the importance of Trp-104 for the activity of IL-3, but not for its interaction with the alpha  receptor, was also clearly identified (27). Therefore, it was suggested that the site defined above actually contained the carbohydrate recognition domain of human IL-3, identified by its strong interaction with the PGS3 proteoglycan.


View larger version (39K):
[in this window]
[in a new window]
  		Fig. 5.   Magnification of the results of docking experiments of the high affinity ligand of IL-3 into the putative carbohydrate recognition domain of IL-3 (A) and computational effect of the mutation R29Q (B). A, the formation of four hydrogen bonds between the IL-3 ligand and Asn-70, His-26 (back), and Arg-29 in the crystallized IL-3. Note that the hydrogen bond of the ligand with Arg-29 in A finds its equivalent with Gln-29 in B, the latter corresponding to the wild-type human IL-3. C, view of the putative CRD of IL-3 showing the van der Waals interactions between the pyranic rings and Trp-104 and His-26 (hydrogen atoms were stripped in order to render the picture clearer). D, proposed binding site of the classic disaccharide of chondroitin-4-sulfate in the P33G mutein.

In Silico Determination of the Three-dimensional Structure of the P33G IL-3 Mutein-- It was previously shown (7, 8) that the P33G mutation provoked 14-fold increased biological activity for the mutein relative to the wild-type recombinant human IL-3. Although the previous authors (7, 8) did not demonstrate strong modification of the three-dimensional structure after the mutation, our in silico studies indicated that the conformation of the mutein (Fig. 4B) was significantly different from that of the non-mutated protein (Fig. 4A). Although the site defined by Bagley et al. (23) as the receptor-binding domain was relatively preserved, the P33G mutation rendered a complete accessibility of Trp-104 and an opening of the cavity in which the IL-3 ligands were initially computationally localized. Especially, the amino acids (His-26, Arg-29, Gln-70) involved in the formation of strong hydrogen bonds with the IL-3 ligand showed a completely modified localization.

Therefore, we performed in silico docking experiments of the five lower energy conformers of the initial IL-3 ligands into the mutein and, for comparison, of the four other disaccharide constituents of chondroitin sulfates mentioned above. In all cases, more than 50% of the lower energy conformers of all disaccharides were localized in the giant cavity of the mutein. However, fine analysis of the interaction of the lower energy conformers of GlcA(2S)beta 1,3GalNAc(4S)beta 1- with the calculated structure of the P33G mutein showed insufficient van der Waals interactions and hydrogen bonding for consideration of this area as a putative CRD for this compound. All the lower energy conformers of GlcA(2S)beta 1,3GalNAc(4S)beta 1 present in this cavity showed very weak van der Waals interactions with Trp-104 (more than 10 Å) and a maximum of one hydrogen bond with surrounding amino acids. When performing the fine analysis of the results of the docking experiments mentioned above, only the classic disaccharide from chondroitin-4-sulfate (GlcAbeta 1,3GalNAc(4S)beta 1-) showed 50% of the lower energy disaccharide/mutein complexes localized close to Trp-104 (Fig. 6). Some of these complexes showed interesting features, such as relatively high van der Waals interactions between the pyranic ring of the 4-sulfated GalNAc and Trp-104 and two hydrogen bonds between the oxygen atom of the sulfate group and the carboxyl group of Asp-103 (Figs. 4B and 5D). Furthermore, additional van der Waals interactions could be possible with Pro-31 and Phe-107, with the three cyclic amino acids constituting a cradle for the disaccharide of chondroitin-4-sulfate. Consequently, these data suggested that in the P33G mutein the putative calculated CRD specific for GlcA(2S)beta 1,3GalNAc(4S)beta 1 changed to a site unable to recognize the previous ligand of IL-3. In contrast, it was suggested that it could recognize, but with a lower affinity, the classic disaccharides of chondroitin-4-sulfate, GlcAbeta 1,3GalNAc(4S)beta 1.


View larger version (40K):
[in this window]
[in a new window]
  		Fig. 6.   Hypothetical schematic drawing explaining the increased activity of the P33G mutein of IL-3. A, the wild-type IL-3 (black) bound to IL-3Ralpha associates the latter to a signal-transducing complex associated with a chondroitin sulfate proteoglycan (purple) containing the rare GlcA(2S)beta 1,3GalNAc(4S)beta 1- disaccharide units, thus generating a signal (s). The majority of the signal-transducing complexes associated with the same chondroitin sulfate containing the abundant GlcAbeta 1,3GalNAc(4S)beta 1- disaccharide units are not able to generate a signal. B, the situation is completely reversed, and the mutein (mIL-3, red) associates IL-3Ralpha to several signal-transducing complexes, generating increased signaling and, consequently, an increased biological response.


    DISCUSSION
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES

Specificity of IL-3 for the GlcA(2S)beta 1,3GalNAc(4S)beta 1 Disaccharide-- Several studies have reported the interactions of cytokines with glycosaminoglycans (3-5). However, using a new procedure involving unlabeled cytokines (1), we were unable to demonstrate such specific interactions with commercially available GAGs because this method (analysis of the quantity of unbound cytokines to immobilized ligand by immunoblotting) was only able to reveal stoichiometric interactions with the immobilized ligand. Furthermore, a significant binding could be detected safely only if at least 10% of the added cytokine were bound. Such a situation never occurred for five cytokines (1) and for seven other cytokines (this study),4 although bindings to commercially available GAGs were detected using labeled cytokines or ELISA techniques (28-31). To analyze these different behaviors, we decided to study the binding of cytokines to proteoglycans isolated from the rat brain (6). This demonstrated that IL-3 recognized, specifically, throughout the different proteoglycans isolated from adult rat brain, the chondroitin sulfate proteoglycan PGS3. This low Mr proteoglycan (Mr 100,000 obtained by gel filtration; Ref. 6) differed from the others by its higher degree of sulfation. Although the oligosaccharide composition of PGS3 was complex (especially in the presence of a significant amount of mannose), the binding of IL-3 to this compound was directed to its GAG portion. In fact, all the mannose of PGS3 was recovered in the dialyzable fraction after beta -elimination, indicating that it was O-linked. Such oligosaccharides were frequently identified in brain proteoglycans (32-36). However, the dialyzable fraction containing these compounds did not show a strong ability for inhibition of the binding of IL-3 to PGS3. In contrast, the interaction was stoichiometrically inhibited by the GAG portion isolated from PGS3.

The higher degree of sulfation was likely necessary for the binding of IL-3 because the majority of the other brain chondroitin-sulfate proteoglycans and commercially available glycosaminoglycans did not show any affinity for IL-3. 1H NMR studies indicated that the particularity of PGS3 consisted in the abundance (2:3) of the disaccharide units GlcA(2S)beta 1,3GalNAc(4S)beta 1, a relatively rare disaccharide previously identified in dog mastocytoma (37).

Definition of the Putative CRD of IL-3-- The computational docking experiments of the lower energy conformers of five different disaccharides derived from chondroitin sulfates in the three-dimensional structure of IL-3 (38, 39) sustained the data obtained by our binding and 1H NMR studies of the GAG portion of PGS3. Indeed, intimate binding was only obtained for GlcA(2S)beta 1,3GalNAc(4S)beta 1 and not for the other compounds. However, it should be stressed that the three-dimensional structure of IL-3 accessible in the literature was not that of the wild-type human IL-3 but a mutated protein (38, 39), in contrast to the IL-3 used here in binding experiments. In fact, the recombinant protein used for the three-dimensional structure determination (38, 39) lacked the 1-13 N-terminal amino acids and the C-terminal residues (126-133). These deletions of the N- and C-terminal portions of IL-3 did not introduce modifications of the receptor binding nor of biological activity on cell cultures. Furthermore, several mutations were performed (V14A, N18I, T25H, Q29R, L32N, F37P, G42S, Q45M, N51R, R55T, E59L, N62V, S67H, and Q69E). Nevertheless, this mutated protein was shown to be identical to the wild-type IL-3, both for its receptor binding and its biological activity on the same cells (38). Therefore, it was suggested that the three-dimensional structure used for the docking experiments was very close to that of the wild-type IL-3 used by us in binding experiments. At least, essential structural features necessary for the full expression of the biological activity of IL-3 and for its receptor binding were preserved. In fact, all these mutations relative to the wild-type were mutations of surface amino acids not involved in the stabilization of the three-dimensional structure of IL-3. Except for the Q29R mutation, all were concerned with the receptor-binding domain and with the putative CRD of IL-3 calculated here. The reason why the Q29R mutation did not provoke modifications of the biological activity of IL-3 was explained by computational calculation of the docking of GlcA(2S)beta 1,3GalNAc(4S)beta 1- in the Q29R mutein. As shown in Fig. 6B, these calculations indicated that the more stable conformation of the complex between the ligand and IL-3 involved the same pattern of van der Waals interactions and hydrogen bonds. The hydrogen bond between the oxygen atom of the carboxylic group of GlcA and the nitrogen atom of the amide group of the lateral chain of Gln-29 was equivalent to that between the same carboxyl group and the nitrogen atom of the guanidinium group of Arg-29. Therefore, it was suggested that the three-dimensional structure of IL-3 reported in the literature constituted a correct model structure for our docking experiments.

The site defined here as a putative CRD was essentially different from those proposed as the receptor-binding domains. Indeed, site-directed mutagenesis (7, 8, 23, 27, 40-43) of human IL-3 defined essential amino acids for binding to the IL-6Ralpha and -3Rbeta as residues Ser-17, Asn-18, Asp-21, Arg-22, Thr-25, Ser-42, Glu-43, Gln-45, Asp-46, Met-49, Arg-94, Pro-96, Arg-108, Lys-110, Leu-111, Phe-113, Lys-116, and Glu-119. Deletion of the first 14 amino acids (40) did not modify the receptor binding of IL-3, a property lost when the first 18 amino acids were deleted (8). No biological activity was found after deletion of the 22 C-terminal amino acids (8). Other mutations (27) defined another site involving residues Trp-104, Asn-105, Lys-110, and Arg-108 as necessary for the activity but not for receptor binding.

Unfortunately, extensive mutagenesis studies were not performed in the site that we propose as the CRD of IL-3 with the exception of Trp-104, which suppressed the activity of IL-3, and the mutation Q29R, which did not change the biological activity of IL-3. This was in agreement with our own computational data on the interaction of the high affinity ligands of IL-3 with these residues. Therefore, our biochemical and docking experiments propose that the second site of IL-3, not necessary for the binding to the IL-3Ralpha but for the biological activity of IL-3, corresponds to its carbohydrate recognition domain. It involves a strong van der Waals interaction between the pyranic ring of GalNAc and Trp-104, a weak van der Waals interaction with His-26, and four strong hydrogen bonds with His-26, (Arg/Gln)-29, and Asn-70. The configuration of this site fulfilled the common features observed for the CRDs of calcium-independent lectins (24-26). This putative CRD of IL-3 is also very similar to that calculated for the CRD of IL-6 in interaction with its HNK-1 type oligosaccharide ligand (2).

Calculated Changes of the Carbohydrate Specificity of the P33G Mutein-- Mutations close to the putative carbohydrate recognition domain defined above (P33N and P33G) increased the biological activity of IL-3 without significant effect on IL-3Ralpha binding (7, 39). Especially, the P33G mutation increased the activity of IL-3 by a factor of 14 without significant changes of the binding to IL-3Ralpha (8). Based on the computer software now available, the authors concluded that this mutation did not provoke essential modifications of the three-dimensional structure of IL-3. Using the Sybyl software, it was evident that the structure of IL-3 was fundamentally changed, with an opening of the molecule and the complete solvent accessibility of the Trp-104 residue. This initially suggested to us a stronger interaction between IL-3 and its ligand, a property that could explain the increased biological activity of the P33G mutein. In fact, the docking experiments contradicted this view, because they indicated that the ligand of IL-3 did not find a significant binding in the region of the putative CRD determined above. In contrast, some lower energy conformers of the disaccharide derived from chondroitin-4-sulfate showed rather good interactions with Trp-104 and the surrounding amino acids. Therefore, it was suggested that the increased biological activity of the P33G mutein could be related to a change in the nature of its chondroitin sulfate ligand.

Indeed, as determined previously for IL-2 (44) and for IL-6 (2), the function of the CRD of the cytokine is the association of its receptor complexes with another surface molecular complex containing the oligosaccharide ligands. This specific extracellular association, due to the bifunctional cytokine, induces specific phosphorylation/dephosphorylation mechanisms (45). A change in specificity of oligosaccharide ligands could induce a change in biological activity without modification of the binding to the receptor. In the case of IL-3, the change from a rare chondroitin sulfate structure (for the wild-type IL-3) to an abundant structure (for the P33G) mutein could produce an increased number of interactions and, consequently, a more intense signaling (Fig. 6). It is worth mentioning that eosinophils and mast cells (or derived cell lines), which are extremely sensitive to IL-3 and were used for testing the effect of mutations on the biological activity of IL-3, contain a significant proportion (6-16%) of disulfated disaccharides (46-48), the remainder being essentially constituted of chondroitin-4-sulfate.

The nature of the proteoglycan ligands at the surface of cells stimulated by IL-3 remains a question. It is commonly accepted that the IL-3 binding to its receptor is responsible for the association of the latter with the common beta -chain of the receptor (23, 49). Based on the absence of consensus polypeptide sequences for the biosynthesis of chondroitin sulfate chains on the IL-3Rbeta , the latter could not be the ligand. Therefore, it is suggested that this beta -chain could be part of a molecular complex containing a highly sulfated chondroitin sulfate proteoglycan, allowing indirect contact between IL-3Ralpha and -3Rbeta . Furthermore, from GC/MS analysis, PGS3 has very short GAG chains (15 disaccharide units), whereas classic chondroitin sulfate proteoglycans (including those of the other brain proteoglycans) have longer GAG chains (more than 100 disaccharide units).5 Therefore, it could be suggested that the increased activity of the P33G mutein could be related to the fact that the increased chain length of the ligand of the mutein can favor the extracellular associations between IL-3Ralpha receptor complexes (carbohydrate-mediated oligomerization) to induce new signaling pathways.

The biological relevance of our findings is supported by the fact that cells sensitive to IL-3, as mentioned above, possess a relatively high level of disulfated chondroitin sulfate chains (although the higher affinity ligand proposed here has not yet been identified in these cells). A further argument comes from the brain immunolocalization of PGS3 (50).6 Indeed, this compound presents a preferential localization on a few neurons of the cerebellum and the forebrain. This localization has to be related to the RT-PCR and immunohistochemical studies showing the presence of IL-3Ralpha in the central nervous system, with a preferential localization in large cholinergic neurons (51-54). Because it has been demonstrated that IL-3 has a neurotrophic effect on these neurons (increases in acetylcholinesterase and choline-acetyltransferase), it may be speculated that the IL-3 signaling occurs through a similar mechanism to that described for the other interleukins (2, 44), i.e. extracellular association of the IL-3Ralpha with a glycoconjugate ligand (here, a chondroitin sulfate proteoglycan) due to the bifunctional IL-3, allowing specific intracellular phosphorylation/dephosphorylation mechanisms.

    FOOTNOTES

* This work was sustained in part by a grant from the Eramus program (to R. B.).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.: 33-03-20-43-40-10; Fax: 33-03-20-43-65-55; E-mail: Jean-Pierre.Zanetta@univ-lille1.fr.

Published, JBC Papers in Press, July 1, 2002, DOI 10.1074/jbc.M205282200

2 J. P. Zanetta, unpublished data.

3 V. Durier, P. Lagant, S. Kirillova, and G. Vergoten, submitted for publication.

4 J. P. Zanetta, unpublished data.

5 J. P. Zanetta, unpublished results.

6 J. P. Zanetta, unpublished results.

    ABBREVIATIONS

The abbreviations used are: CRD, carbohydrate recognition domain; BCIP, 5-bromo-4-chloro-3-indolyl-phosphate; PBS, phosphate-buffered saline; CPC, cetylpyridinium chloride; GAG, glycosaminoglycan; GlcA, D-glucuronic acid; GalNAc, D-N-acetylgalactosamine; GC/MS, gas chromatography/mass spectrometry; IL, interleukin; NBT, nitro blue tetrazolium; GC/MS, gas chromatography/mass spectrometry; ELISA, enzyme-linked immunosorbent assay.

    REFERENCES
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES

1. Cebo, C., Dambrouck, T., Maes, E., Laden, C., Strecker, G., Michalski, J.-C., and Zanetta, J. P. (2001) J. Biol. Chem. 276, 5685-5691[Abstract/Free Full Text]
2. Cebo, C., Durier, V., Lagant, P., Maes, E., Florea, D., Lefebvre, T., Strecker, G., Vergoten, G., and Zanetta, J. P. (2002) J. Biol. Chem. 277, 12246-12252[Abstract/Free Full Text]
3. Roberts, R., Gallagher, J., Spooncer, E., Allen, T. D., Bloomfield, F., and Dexter, T. M. (1988) Nature 332, 376-378[CrossRef][Medline] [Order article via Infotrieve]
4. Kolset, S. O., and Gallagher, J. T. (1990) Biochim. Biophys. Acta 1032, 191-211[Medline] [Order article via Infotrieve]
5. Gupta, P., Oegema, T. R., Jr., Brazil, J. J., Dudek, A. Z., Slungaard, A., and Verfaillie, C. M. (2000) Blood 95, 147-155[Abstract/Free Full Text]
6. Normand, G., Kuchler, S., Meyer, A., Vincendon, G., and Zanetta, J. P. (1988) J. Neurochem. 51, 665-676[CrossRef][Medline] [Order article via Infotrieve]
7. Lokker, N. A., Movva, N. R., Strittmatter, U., Fagg, B., and Zenke, G. (1991) J. Biol. Chem. 266, 10624-10631[Abstract/Free Full Text]
8. Lokker, N. A., Zenke, G., Strittmatter, U., Fagg, B., and Movva, N. R. (1991) EMBO J. 10, 2125-2131[Medline] [Order article via Infotrieve]
9. Glass, W. F., II, Briggs, R. C., and Hnilica, L. S. (1981) Anal. Biochem. 115, 219-224[CrossRef][Medline] [Order article via Infotrieve]
10. Suzuki, K. (1964) Life Sci. 3, 1227-1233[CrossRef][Medline] [Order article via Infotrieve]
11. Endo, M., and Yosizawa, Z. (1979) Tohoku J. Exp. Med. 129, 225-231[Medline] [Order article via Infotrieve]
12. Brunngraber, E. G., Brown, B. D., and Aguilar, V. (1969) J. Neurochem. 16, 1059-1070[CrossRef][Medline] [Order article via Infotrieve]
13. Dische, Z. (1947) J. Biol. Chem. 167, 189-198[Free Full Text]
14. Zanetta, J. P., Timmerman, P., and Leroy, Y. (1999) Glycoconj. J. 16, 617-627[CrossRef][Medline] [Order article via Infotrieve]
15. Laemmli, U. K. (1970) Nature 227, 680-685[CrossRef][Medline] [Order article via Infotrieve]
16. Towbin, H., Staehelin, T., and Gordon, J. (1979) Proc. Natl. Acad. Sci. U. S. A. 76, 4350-4354[Abstract/Free Full Text]
17. Leary, J. J., Brigati, D. J., and Ward, D. C. (1983) Proc. Natl. Acad. Sci. U. S. A. 80, 4045-4049[Abstract/Free Full Text]
18. Katchalski-Katzir, E., Shariv, I., Eisenstein, M., Friesen, A. A., Aflalo, C., and Vasker, I. A. (1992) Proc. Natl. Acad. Sci. U. S. A. 89, 2195-2199[Abstract/Free Full Text]
19. Clement, A. M., Nadanaka, S., Masayama, K., Mandl, C., Sugahara, K., and Faissner, A. (1998) J. Biol. Chem. 273, 28444-28453[Abstract/Free Full Text]
20. Nadanaka, S., Clement, A., Masayama, K., Faissner, A., and Sugahara, K. (1998) J. Biol. Chem. 273, 3296-3307[Abstract/Free Full Text]
21. Gustin, S. E., Church, A. P., Ford, S. C., Mann, D. A., Carr, P. D., Ollis, D. L., and Young, I. G. (2001) Eur. J. Biochem. 268, 2905-2911[Medline] [Order article via Infotrieve]
22. Kinoshita, A., Yamada, S., Haslam, S. M., Morris, H. R., Dell, A., and Sugahara, K. (2001) Biochemistry 40, 12654-12665[CrossRef][Medline] [Order article via Infotrieve]
23. Bagley, C. J., Phillips, J., Cambareri, B., Vadas, M. A., and Lopez, A. F. (1996) J. Biol. Chem. 271, 31922-31928[Abstract/Free Full Text]
24. Lobsanov, Y. D., Gitt, M. A., Leffler, H., Barondes, S. H., and Rini, J. M. (1993) J. Biol. Chem. 268, 27034-27038[Abstract/Free Full Text]
25. Dodd, R. B., and Drickamer, K. (2001) Glycobiology 11, 71-79[CrossRef]
26. Drickamer, K. (1997) Structure 5, 465-468[Medline] [Order article via Infotrieve]
27. Lopez, A. F., Shannon, M. F., Barry, S., Phillips, J. A., Cambareri, B., Dottore, M., Simmons, P., and Vadas, M. A. (1992) Proc. Natl. Acad. Sci. U. S. A. 89, 11842-11846[Abstract/Free Full Text]
28. Kuschert, G. S. V., Hoogewerf, A. J., Proudfoot, A. E. I., Chung, C. W., Cooke, R. M., Hubbard, R., E., Wells, T. N. C., and Sanderson, P. N. (1998) Biochemistry 37, 11193-11201[CrossRef][Medline] [Order article via Infotrieve]
29. Clarke, D., Katoh, O., Gibbs, R. V., Griffiths, S. D., and Gordon, M. Y. (1995) Cytokine 7, 325-330[CrossRef][Medline] [Order article via Infotrieve]
30. Lortat-Jacob, H., Garrone, P., Banchereau, J., and Grimaud, J. A. (1997) Cytokine 9, 101-105[CrossRef][Medline] [Order article via Infotrieve]
31. Najjam, S., Gibbs, R., Gordon, M. Y., and Rider, C. C. (1997) Cytokine 9, 1013-1022[CrossRef][Medline] [Order article via Infotrieve]
32. Finne, J., Krusius, T., Margolis, R. K., and Margolis, R. U. (1979) J. Biol. Chem. 254, 10295-10300[Abstract/Free Full Text]
33. Krusius, T., Finne, J., Margolis, R. K., and Margolis, R. U. (1986) J. Biol. Chem. 261, 8237-8242[Abstract/Free Full Text]
34. Krusius, T., Reinhold, V. N., Margolis, R. K., and Margolis, R. U. (1987) Biochem. J. 245, 229-234[Medline] [Order article via Infotrieve]
35. Smalheiser, N. R., Haslam, S. M., Sutton-Smith, M., Morris, H. R., and Dell, A. (1996) J. Biol. Chem. 273, 23698-23703
36. Smalheiser, N. R., and Kim, E. (1995) J. Biol. Chem. 270, 15425-15433[Abstract/Free Full Text]
37. Forsberg, L. S., Lazarus, S. C., Seno, N., DeVinney, R., Caughey, G. H., and Gold, W. M. (1988) Biochim. Biophys. Acta 967, 416-428[Medline] [Order article via Infotrieve]
38. Feng, Y., Klein, B. K., and McWherter, C. A. (1996) J. Mol. Biol. 259, 524-541[CrossRef][Medline] [Order article via Infotrieve]
39. Feng, Y., Hood, W. F., Forgey, R. W., Abegg, A. L., Caparon, M. H., Thiele, B. R., Leimgruber, R. M., and McWherter, C. A. (1997) Protein Sci. 6, 1777-1782[Medline] [Order article via Infotrieve]
40. Olins, P. O., Bauer, S. C., Braford-Goldberg, S., Sterbenz, K., Polazzi, J. O., Caparon, M. H., Klein, B. K., Easton, A. M., Paik, K., and Klover, J. A. (1995) J. Biol. Chem. 270, 23754-23760[Abstract/Free Full Text]
41. Klein, B. K., Olins, P. O., Bauer, S. C., Caparon, M. H., Easton, A. M., Braford, S. R., Abrams, M. A., Klover, J. A., Paik, K., Thomas, J. W., Hood, W. F., Shieh, J. J., Polazzi, J. O., Donnelly, A. M., Zeng, D. L., Welply, J. K., and McKearn, J. P. (1999) Exp. Hematol. 27, 1746-1756[CrossRef][Medline] [Order article via Infotrieve]
42. Barry, S. C., Bagley, C. J., Phillips, J., Dottore, M., Cambareri, B., Moretti, P., D'Andrea, R., Goodall, G. J., Shannon, M. F., and Vadas, M. A. (1994) J. Biol. Chem. 269, 8488-8492[Abstract/Free Full Text]
43. Klein, B. K., Feng, Y., McWherter, C. A., Hood, W. F., Paik, K., and McKearn, J. P. (1997) J. Biol. Chem. 272, 22630-22641[Abstract/Free Full Text]
44. Zanetta, J. P., Alonso, C., and Michalski, J. C. (1996) Biochem. J. 318, 49-53[Medline] [Order article via Infotrieve]
45. Chin, H., Saito, T., Arai, A., Yamamoto, K., Kamiyama, R., Miyasaka, N., and Miura, O. (1997) Biochem. Biophys. Res. Commun. 239, 412-417[CrossRef][Medline] [Order article via Infotrieve]
46. Rothenberg, M. E., Pomerantz, J. L., Owen, W. F. Jr., Avraham, S., Soberman, R. J., Austen, K. F., and Stevens, R. L. (1988) J. Biol. Chem. 263, 13901-13908[Abstract/Free Full Text]
47. Rothenberg, M. E., and Austen, K. F. (1989) Ann. N. Y. Acad. Sci. 556, 233-244[CrossRef][Medline] [Order article via Infotrieve]
48. Arora, N., Min, K. U., Costa, J. J., Rhim, J. S., and Metcalfe, D. D. (1993) Int. Arch. Allergy Immunol. 100, 319-327[Medline] [Order article via Infotrieve]
49. Bagley, C. J., Woodcock, J. M., Hercus, T. R., Shannon, M. F., and Lopez, A. F. (1995) J. Leukocyte Biol. 57, 739-746[Abstract]
50. Zanetta, J. P., Cebo, C., and Vergoten, G. (2002) Biochim. Biophys. Acta, in press
51. Kamegai, M., Niijima, K., Kunishita, T., Nishizawa, M., Ogawa, M., Araki, M., Ueki, A., Konishi, Y., and Tabira, T. (1990) Neuron 4, 429-436[CrossRef][Medline] [Order article via Infotrieve]
52. Konishi, Y., Chui, D. H., Kunishita, T., Yamamura, T., Higashi, Y., and Tabira, T. (1995) J. Neurosci. Res. 41, 572-582[CrossRef][Medline] [Order article via Infotrieve]
53. Tabira, T., Chui, D. H., Fan, J. P., Shirabe, T., and Konishi, Y. (1998) Ann. N. Y. Acad. Sci. 840, 107-116[CrossRef][Medline] [Order article via Infotrieve]
54. Wen, T. C., Tanaka, J., Peng, H., Desaki, J., Matsuda, S., Maeda, N., Fujita, H., Sato, K., and Sakanaka, M. (1998) J. Exp. Med. 188, 635-649[Abstract/Free Full Text]

Copyright © 2002 by The American Society for Biochemistry and Molecular Biology, Inc.
Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Reddit Reddit   Add to Technorati Technorati    What's this?



This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
277/41/38764    most recent
M205282200v1
Right arrow Submit a Letter to Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Zanetta, J.-P.
Right arrow Articles by Vergoten, G.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Zanetta, J.-P.
Right arrow Articles by Vergoten, G.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
 All ASBMB Journals   Molecular and Cellular Proteomics 
 Journal of Lipid Research   ASBMB Today 
Copyright © 2002 by the American Society for Biochemistry and Molecular Biology.
Advertisement
spacer
Advertisement
Advertisement