Biochimica et Biophysica Acta 1687 (2005) 52 – 63 http://www.elsevier.com/locate/bba
Regular paper
Chemokines bind to sulfatides as revealed by surface plasmon resonance Roger Sandhoff a,*, Heike Grieshabera, Roghieh Djafarzadehb, Tjeerd P. Sijmonsmaa, Amanda E.I. Proudfootc, Tracy M. Handeld, Herbert Wiegandta, Peter J. Nelsonb, Hermann-Josef Grfnea a
German Cancer Research Center, Department of Cellular and Molecular Pathology, INF 280, 69120 Heidelberg, Germany b Medical Policlinic, Ludwig-Maximilians-University of Munich, Germany c Serono Pharmaceutical Institute, Geneva, Switzerland d Department of Molecular and Cell Biology, University of California at Berkeley, Berkeley, CA 94720, USA Received 26 September 2004; received in revised form 5 November 2004; accepted 8 November 2004 Available online 10 December 2004
Abstract Chemokines bind to sulfated cell surface glycosaminoglycans and thereby modulate signaling mediated by G-protein-coupled seventransmembrane domain chemokine receptors. Similar to glycosaminoglycans, sulfated oligosaccharides are also exposed on the cell surface by sulfatides, a class of glycosphingolipids. We have now identified sulfated glycosphingolipids (sulfatides) as novel binding partners for chemokines. Using surface plasmon resonance (SPR), the binding of proinflammatory and homeostatic chemokines to glycosphingolipids, in particular sulfatides, was investigated. Chemokines were immobilized while glycosphingolipids or additional phospholipids incorporated into liposomes were applied as soluble analytes. A specific affinity of the chemokines MCP-1/CCL2, IL-8/CXCL8, SDF-1a/CXCL12, MIP-1a/ CCL3 and MIP-1h/CCL4 to the sulfatides SM4s, SM3, SM2a and SB2, SB1a was detected. No significant interactions with the chemokines were observed for gangliosides, neutral glycosphingolipids or phospholipids. Chemokine receptors have been associated with the detergentinsoluble fraction supposed to contain draftsT, i.e., glycosphingolipid enriched microdomains of the cell surface. Accordingly, the data suggest that early chemokine receptor signaling may take place in the vicinity of sulfated glycosphingolipids on the cell surface, whereby these sulfatides could modulate the chemokine receptor-mediated cell activation signal. D 2004 Elsevier B.V. All rights reserved. Keywords: Chemokine; Sulfatide; Surface plasmon resonance; Liposome; Glycosphingolipid; Glycosaminoglycan
Abbreviations: Chemokine short hand designations were IL-8, interleukin 8; MCP-1, monocyte chemo-attractant protein-1; MIP-1, macrophage inflammatory protein-1; SDF-1a, stromal cell-derived factor; EDC, NV-ethyl-NV-(3-dimethylaminopropyl)carbodiimide; GAG(s), glycosaminoglycans; Ganglioside short hand designations were GM3, II3-N-acetyl(or N-glycolyl)-neuraminyl-lactosylceramide; GM2, II3-N-acetyl(or N-glycolyl)-neuraminylgangliotriaosylceramide; GM1a, II3-N-acetyl(or N-glycolyl)-neuraminyl-gangliotetraosylceramide; GD1a, II3,IV3-bis-N-acetyl(or N-glycolyl)-neuraminylgangliotetraosylceramide; GSL(s), glycosphingolipid(s); glycosphingolipid short hand designations were GlcCer, glucosylceramide; LacCer, lactosylceramide; Gg3Cer, gangliotriaosylceramide; Gg4Cer, gangliotetraosylceramide; NHS, N-hydroxysuccinimide; Phospholipid short hand designations were PE, phosphatidylethanolamine; PC, phosphatidylcholine; PI, phosphatidylinositol; PS, phosphatidylserine; SM, sphingomyelin; [RU], response unit; Sugar short hand designations were Fuc, fucose; Gal, galactose; GalNAc, N-Acetyl-galactosamine; Glc, glucose; GlcNAc, N-acetyl-glucosamine-; GlcA, glucuronic acid; Ins, inositol; Man, mannose; NeuNAc, N-acetyl-neuraminic acid; Sulfatides were abbreviated according to Ishizuka [52], i.e., SM4s, galactosylceramide sulfate, GalCer I3-sulfate; SM3, lactosylceramide sulfate, LacCer II3-sulfate; SM2a, gangliotriaosylceramide II3-sulfate; SB1a, gangliotetraosylceramide II3, IV3-bis-sulfate * Corresponding author. Tel.: +49 6221 424358; fax: +49 6221 424352. E-mail address:
[email protected] (R. Sandhoff). 1388-1981/$ - see front matter D 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.bbalip.2004.11.011
R. Sandhoff et al. / Biochimica et Biophysica Acta 1687 (2005) 52–63
1. Introduction Chemokines are small chemotactic cytokines that induce the adhesion, migration and activation of diverse cell types. The chemokine subfamilies are grouped according to the orientation of specific critical cysteine residues and are designated as CXC (e.g., interleukin-8/CXCL8), CC (e.g., RANTES/CCL5), C (lymphotactin/CL1), or CX3CL (fractalkine/CX3CL-1) chemokines. Chemokines are ligands for a superfamily of G-protein-coupled seven-transmembrane receptors which mediate their biological activities. These receptors are found within the detergent-insoluble fraction at 4 8C thought to be enriched with molecules from lipid rafts [1]. Lipid rafts are discussed to act as signaling platforms on the cell surface. Chemokines possess two major noncovalent binding epitopes: a high affinity site responsible for specific ligand/receptor interactions and a lower affinity site, also called the heparin- or glycosaminoglycan-binding domain, responsible for the presentation of chemokines on the surface of endothelial cells and extracellular matrix. Some studies have shown that glycosaminoglycan (GAG) binding is not essential for the in vitro activity of chemokines, although it can assist the recruitment of the chemokine to the cell surface [2]. However, formal
53
evidence has recently demonstrated that the interaction is relevant for chemokine activity in vivo [3]. Binding of chemokines to heparan sulfate proteoglycans on the surface of endothelial cells also seems to be crucial for the recruitment of leukocytes to inflammatory sites [3–5]. Glycosaminoglycans (GAGs) are long, unbranched polysaccharide chains composed of repeating disaccharide units. In most cases the amino-sugar is sulfated, and with the exception of hyaluronic acid and heparin, all GAGs are covalently bound to proteins in the form of proteoglycans. The amino acid residues responsible for GAG-binding capacity have been characterized for certain chemokines [3,6–13] and reveal different patterns that suggest specificity in the GAG interaction. Sulfated saccharide structures are also found on cellsurface glycosphingolipids. These sulfatides (the general name for all sulfated glycosphingolipids, see Fig. 1 for structures) have been shown to bind diverse proteins including hepatocyte growth factor [14], thrombospondin [15], laminin [16], brevican [17], selectins [18,19], as well as properdin and factor H, regulators of the alternative pathway of complement activation [20]. The specific expression of sulfatides on granulocytes, erythrocytes, platelets and certain tumor cells has also been demonstrated [21–27]. According to Mamelak et al. [28],
Fig. 1. Glycosphingolipid structures. Glycosphingolipids (GSLs) are components of the plasma membrane. Neutral GSLs consist of a lipophilic ceramideanchor and a polar carbohydrate-head group. Acidic GSLs contain either sulfate(s) (sulfatides) or sialic acid(s) (gangliosides) in addition, or both.
54
R. Sandhoff et al. / Biochimica et Biophysica Acta 1687 (2005) 52–63
Hsp70s contain a specific sulfogalactolipid binding site. In addition, the N-terminal carbohydrate recognition domain of galectin-8 recognizes specific glycosphingolipids with high affinity including sulfatides (SM4s), SM3 and SB1a [29]. Sulfatides are located in the extracellular leaflet of the plasma membrane and are thought to be enriched in lipid-rafts where the chemokine-receptors also appear to be present [1]. Because of this distribution, it was thought possible that chemokines may specifically interact with sulfatides at the signaling platform of a cell. An investigation was initiated aiming at the detection of possible glycolipid chemokine interactions by using surface plasmon resonance (SPR). SPR was chosen for its noninvasive technology with no need of compound labeling. It is based on measurement of light reflected from a sensor surface to which one of the interaction partners is immobilized. Surface interaction of a soluble binding partner results in a signal change that is directly proportional to the change in mass. Hence, using a bigger binding partner in soluble form will yield a more sensitive signal. As compared to sulfatides, incorporated into 100-nm liposomes, chemokines are relatively small molecules (molecular weight approximately 8 kDa), and therefore were immobilized. With this approach, we showed that the chemokines MCP-1/CCL2, IL-8/CXCL8, SDF-1a/CXCL12, MIP-1a/ CCL3 and MIP-1h/CCL4 bind selectively to sulfatide structures. Furthermore, we demonstrated that the sulfatide binding overlaps the previously characterized GAG binding site of MCP-1/CCL2.
2. Materials and methods 2.1. Materials All chemicals and solvents were of p.A. grade. SM3 was purified from human kidney and SM2a, SB2, and SB1a were isolated from rat kidney and quantified according to Sandhoff et al. [30]. Sulfatide (SM4s), GlcCer, GalCer, LacCer, GM3, Gg3Cer and bovine brain gangliosides were purchased from Matreya (PA 16823 USA), GM2 was an isolate from human GM2-gangliosidosis brain. Commercially available GSLs or GSLs from tissue, isolated and purified, were quantified with anthrone reaction according to Ref. [31]. Phospholipids (PE, PC, PI and PS) and Cholera toxin B-subunit were obtained from Sigma-Aldrich (Deisenhofen, Germany). Recombinant MCP-1/CCL2, IL-8/CXCL8 (77AA), SDF-1a/CXCL12, MIP-1a/CCL3, and MIP-1h/CCL4 were bought from PreproTech (London, Great Britain). The MCP-1/CCL2 mutant (18AA19-MCP-1/CCL2) with point mutations R18A and K19A in a background containing the silent mutation M64I (which has no affect on function) has been recently described [3]. Recombinant MCP-1/CCL2 containing six additional His residues at the C-terminus was available from Serono Laboratories.
2.2. Methods 2.2.1. Liposome preparation Liposomes were prepared at a total concentration of 1.5 mM and diluted with buffer according to the experimental conditions. Lipids dissolved in chloroform/methanol were mixed according to their molar percentage in the corresponding liposomes. The solvent was evaporated under a gentle nitrogen stream. The lipids were dissolved in ethanol and again dried. The dry lipid mixture was then resuspended with 500-Al buffer, and incubated for 1 min in an ultrasonic bath. Subsequently, the samples were subjected to 10 cycles of freezing and melting with liquid nitrogen and a 37 8C warm water bath. Suspensions containing the liposomes were extruded 21 times through 100-nm polycarbonate membranes in a Liposofast apparatus (Avestin, Ottawa, ON, Canada). Liposomes were separated from aggregates by a 5min centrifugation step in a bench top centrifuge at 20 800g. Liposome preparations were stored at 4 8C in 10 mM Tris buffer, pH 7.4, containing 150 mM NaCl. 2.2.1.1. Size measurement of liposomes. Liposomes were measured six times for 1 min by the dynamic light scattering technique using a zetasizer 1000 HSA (Malvern, Worcestershire, UK). 2.2.2. SPR analysis Binding experiments were performed by SPR using a BIAcore 2000TM biosensor system (Biacore, Piscataway, NJ). Three types of chips were used, the lipophilic surface L1 chip binding lipid material, the CM5 chip that allows carbodiimide-coupling via free primary amino-groups of the ligand, and the NTA chip immobilizing peptides containing a His4–6-sequence (His-tag) via a Ni-complex (Biacore). To establish the SPR protocol for further experiments with GSL and chemokines, binding of ganglioside GM1 to cholera toxin B-subunit (CTX-B5), one of the highest affinity complexes known, was used as a model [32–34]. GM1 was inserted into liposomes (5 mol% GM1, 75 mol% phosphatidylcholine (PC) and 20 mol% cholesterol), and immobilized on a liposome-binding L1 chip surface. Control-liposomes containing no GM1, but 5% GM3 or additional 5% PC, were loaded on other flow cells of the same L1 chip. Cholera toxin B-subunit was injected at a concentration of 10 Ag/ml. For inverse configuration, Cholera toxin B-subunit was immobilized on a CM5 sensor chip-surface and liposomes were injected. 2.2.2.1. CM5 chip. Prior to chemokine immobilization, the CM5 sensor chip-surface was activated with a mixture of 0.4 M EDC and 0.1 M NHS (flow rate 2 Al/min, injection volume 20 Al). The chemokines SDF-1a/CXCL12, IL-8/ CXCL8, MCP-1/CCL2, and its non-GAG-binding mutant were immobilized on CM5 sensor chips at a concentration of 20 Ag/ml in distilled water, whereas MIP-1a/CCL3 and MIP-1h/CCL4 (20 Ag/ml) were injected in 10 mM acetate
R. Sandhoff et al. / Biochimica et Biophysica Acta 1687 (2005) 52–63
pH 4 (flow rate 2 Al/min). The injection time (0.5 to 5 min) varied depending on the desired immobilization rate (300– 3000 RU). Analyses were performed at 25 8C and a flow rate of 20 Al/min. Liposomes were injected at a physiological pH using a buffer of Tris/HCl pH 7.4 containing 150 mM NaCl for 2 min with a flow rate of 20 Al/min. They were composed of 55% phosphatidylcholine (PC), 20% cholesterol and 25 mol% of either sulfatide (SM4s, SM3, SM2a, SB2, or SB1a), ganglioside (GM3 GM2, GM1a, or GD1a), neutral glycosphingolipids (GalCer, LacCer, or Gg4Cer), or additional phospholipids [phosphatidylcholine (PC), phosphatidylethanolamine (PE), phosphatidylserine (PS) or phosphatidylinositol (PI)] at a total lipid concentration in the medium of 150 AM (corresponding to an estimated concentration of 19 AM glycolipid on the outside of the liposomes). At the end of the injection, the CM5 chip surface was regenerated by washing off bound liposomes with a solution of 40 mM h-octylglucoside for 1 min. 2.2.2.2. L1 chip. Liposomes with a total lipid concentration in the aqueous medium of 1.5 mM were injected at a flow rate of 2 Al/min onto the corresponding flow cell surface of the L1 chip. If a loading status of 7000–10 000 RU could not be achieved with prolonged injection (which was the case for 25 mol% sulfatide containing liposomes), the flow cell was blocked with a second injection of cholesterol/PCliposomes. Nonspecifically bound liposomes were subsequently washed off with a 1-min pulse of 0.1 M NaOH at a flow rate of 10 Al/min. Proteins that remained bound to the immobilized liposomes after the end of injection were washed off with 4 M MgCl2. The L1 chip surface was regenerated from bound liposomes by washing with 40 mM h-octylglucoside. 2.2.2.3. NTA chip. Immobilization of His-tagged proteins was carried out according to the company’s protocol. In brief, a running buffer containing 10 mM HEPES, 150 mM NaCl, 50 AM EDTA, and 0.005% Tween 20, pH 7.4 was used. The chip surface was washed with a regeneration solution containing 350 mM EDTA in running buffer (20 Al/ min) and activated with a Nickel-solution containing 500 AM NiCl2 in running buffer (5 Al/min). His-tagged proteins at a concentration of 20 Ag/ml in running buffer were immobilized at a flow rate of 5 Al/min. Liposomes (150 AM total lipid concentration) were injected for 2 1/2 min at a flow rate of 10 Al/min.
3. Results 3.1. SRP measurement of ganglioside GM1 binding to cholera toxin The interaction between cholera toxin B-subunit and the ganglioside GM1 has previously been measured with different assays including SPR. It is one of the strongest
55
ligand protein interactions known, with reported K D values determined by SPR-analysis in the range from 2.610 10 to 4.6110 12 M [32–34]. In good agreement with the earlier reports, in the present investigation, a K D value of 7.610 11 for the GM1–CTX-B5 interaction was determined for immobilized lipid and CTX-B5 as the analyte (data not shown). When CTX-B5 was immobilized on a CM5-sensor chip surface and ganglioside GM1-containing liposomes were used as analyte, a threefold higher response signal as compared to the reverse application, i.e., immobilized GM1-liposomes and solute cholera toxin B-subunit, was obtained. No detachment of the analyte (liposomes) was detected after injection stop (data not shown). Because of this increase in response signal, it was decided to measure chemokine–glycolipid interactions with the latter approach. 3.2. Sulfatide binding to MCP-1/CCL2 MCP-1/CCL2 covalently linked to a CM5-sensor chip surface showed binding to liposomes that contained the sulfatides SM4s, SM3, SB2, or SB1a. Almost no dissociation of the liposomes after binding to the chemokines was observed. Of all sulfatides investigated, SM3 showed the greatest capacity for chemokine binding when divided by the number of sulfate groups per molecule (SM4s and SM3: one sulfate group; SB2 and SB1a: two sulfate groups) (Fig. 2). Liposomal sizes were determined for three neutral GSLs and their corresponding sialylated and sulfated derivatives. The different liposomes varied between 90 and 170 nm, with liposomes containing neutral lipids being on average smaller than those with gangliosides or sulfatides. Within this range, liposomal size did not have a detectable influence on binding specificity to MCP-1/ CCL2. For example, ganglioside-containing liposomes had the same or even larger size as corresponding sulfatide-containing liposomes, but, in contrast to the sulfatides, did not bind to MCP-1/CCL2 (Table 1). Binding of liposomes containing either SM4s, SM3, SB2, or SM2a to immobilized MCP-1/CCL2 was independent of Ca2+ (data not shown). In addition, the SRP measurements of the sulfatide-liposome binding to immobilized chemokines showed no obvious dependence on the pH in the range of 5.9 to 8.2 (data not shown). Only in the case of SB2-liposomes, binding to chemokines was reduced to less than 50% at pH 8.2. However, the interaction between sulfatide and chemokines is dependent on the concentration of both (Fig. 3A–D). Immobilizing only half the amount of MCP-1/CCL2 reduced the amount of sulfatide bound to approximately 10% (SM3) or less (SB2, SM2 and SM4) (Fig. 3B). This nonlinear concentration dependency was not observed for SDF-1a/CXCL12. Reduction of SDF-1a/CXCL12 by 40% of its concentration on the chip surface resulted in a roughly linear decrease of sulfatide binding down to 25–40%.
56
R. Sandhoff et al. / Biochimica et Biophysica Acta 1687 (2005) 52–63
Fig. 2. SPR sensogram: immobilized chemokine MCP-1/CCL2 interacts specifically with sulfatides containing liposomes. MCP-1/CCL2 was immobilized onto a CM5-sensor chip surface with 2990 RU. Liposomes containing 20 mol% cholesterol, 55 mol% PC and 25 mol% of the indicated Lipid were injected (start: 0 s) for 150 s in a liposomal lipid concentration of 150 AM and at a flow rate of 20 Al/min [indicated lipid: PC (light green), PE (green), GalCer (light blue), LacCer (light cyan), Gg4Cer (cyan), GM2 (gray), GM1 (light gray), GD1a (gray), PI (yellow), PS (light yellow), SM4s (dark red), SM3 (dark blue), SM2a (light magenta), SB2 (light red) and SB1a (magenta)]. Running and injection buffer: 50 mM Tris, 150 mM NaCl, pH 7.4. Spikes at 0 and 150 s are due to the pumps and IFC system.
3.3. HIS-tagged MCP-1/CCL2 immobilized via a Ni-complex shows similar binding specificity In order to investigate the possible influence of coupling the chemokines to the chip matrix via free amino groups, a recombinant MCP-1/CCL2 containing six additional His residues at the C-terminus was immobilized onto a Niactivated NTA chip. The immobilization was not stable (similar to other his-tagged proteins [35]) and slow bleeding of the protein during the experiments occurred (from initially 100% bound protein to about 80%). The selective interaction of chemokine coupled in this way to liposomes containing sulfatides SM4s, SM3, SB2, and SB1a was still clearly observed (Fig. 4). Similar to the results seen with chemokines coupled via amino-groups, the sulfatide SM3
Table 1 Correlation of liposomal size and strength of binding to MCP-1/CCL2 by SPR technique Liposomes containing 25 mol% of the following lipid
Liposomal size [nm]
Binding to MCP-1/ CCL2 [RU]
PC LacCer GM3 SM3 Gg3Cer GM2 SM2 Gg4Cer GD1 SB1a
91F12a 113F23 161F19 158F20 104F25 173F14 131F16 116F27 141F22 156F4
58F35b 18F23 23F2 1490F16 111F45 27F10 318F45 60F12 2F6 1301F72
a b
Mean of six measurementsFstandard deviation. Mean of two measurementsFrange.
displayed the greatest binding capacity in relation to the number of sulfate residues present. Thus, coupling of the chemokines via their amino-groups does not greatly influence sulfatide binding capacity, at least for MCP-1/ CCL2. 3.4. The GAG-binding residues participate in sulfatide binding of MCP-1/CCL2 The specific amino acid residues important for chemokine binding to GAGs have been identified for a series of chemokines. For MCP-1/CCL2, the double mutation R18A,K19A results in a dramatic reduction in GAG binding [3]. To determine whether or not this region plays a role in binding to sulfatides, wild-type MCP-1/CCL2 and a mutant 18 AA19-MCP-1/CCL2 (where alanine residues replace arginine and lysine) were immobilized onto a CM5 chip surface. As was the case for wild-type MCP-1/CCL2, the mutant showed no binding to control-liposomes. Interaction with the sulfatide-containing liposomes decreased significantly with the mutant chemokine in comparison to wild type (Fig. 5). The interaction with the monosulfo-sulfatides was reduced by approximately half (30–60% binding for SM4s, 20–50% for SM3) and even more for the bis-sulfosulfatides (3–9% binding for SB2 or 9–20% for SB1a). 3.5. Chemokines bind with varying specificities to complex sulfatides In addition to MCP-1/CCL2, the chemokines IL-8/ CXCL8, SDF-1a/CXCL12, MIP-1h/CCL4, and MIP-1a/ CCL3 were covalently linked to the CM5-sensor chip
R. Sandhoff et al. / Biochimica et Biophysica Acta 1687 (2005) 52–63
57
Fig. 3. Dependency of the liposome–chemokine interaction on the concentration of immobilized MCP-1/CCL2 or SDF-1a/CXCL12 or liposomes in solution. (A and C) Liposomes containing 25% PC, SM4s, SM3, SB2, SM2a or GM2 were injected at different liposomal lipid concentrations (MCP-1/CCL2: 1500, 150, 15, and 1.5 AM lipids; SDF-1a/CXCL12: 150, 75, 37.5, and 18.75 AM) onto a (A) MCP-1/CCL2 (1925 RU)- or (C) SDF-1a/CXCL12 (1800 RU)covered CM5 chip surface, each for 150 s. (B) MCP-1/CCL2 and (D) SDF-1a/CXCL12 were immobilized onto a CM5 chip surface in three different concentrations (MCP-1/CCL2: 465, 957, and 1925 RU; SDF-1a/CXCL12: 750, 1500 and 1800 RU). Liposomes containing 25% PC, SM4s, SM3, SB2, SM2a or GM2 were injected at a liposomal lipid concentration of 1500 AM (MCP-1/CCL2) and 150 AM (SDF-1a/CXCL12) for 150 s. Relative response [RU] represents the amount of bound liposomes after 145-s injection time. Bars represent the mean of two measurements. The data range is depicted by the lines on top of the bars.
surface via amino-groups at a concentration of 300–2800 RU. Liposomes containing 25 mol% sulfatide (SM4s, SM3, SM2a, SB2, or SB1a), ganglioside (GM3, GM2, GM1a, or GD1a), LacCer or additional phospholipid (PC) were
injected at a lipid-concentration of 150 AM. All chemokines bound liposomes containing sulfatides, however, with greatly varying specificities. Thus, within the group of sulfatides a differential binding was observed for specific
Fig. 4. SPR sensogram of HIS-tagged MCP-1/CCL2 with sulfatide-containing liposomes. Recombinant MCP-1/CCL2 with six additional histidines at the Cterminus was immobilized onto an NTA-sensor chip surface through the interaction of the bHIS-TAGQ with the chelated Ni-ion on the chip surface. Liposomes containing 20 mol% cholesterol, 55 mol% PC and 25 mol% PC (light green), SM4s (dark red), SM3 (dark blue), SM2a (light magenta), SB2 (light red), SB1a (magenta), or GM2 (gray) were injected for 150 s in a liposomal lipid concentration of 150 AM. Running and injection buffer: 50 mM HEPES, 150 mM NaCl, 50 AM EDTA, 0.005% Tween 20, pH 7.4. Flow rate: 10 Al/min. Since the MCP-1/CCL2-HIS6 slowly bleeds off the chip surface, the loading status at injection time 0 of each liposome-injection was measured: (a) SM3: 3960 RU, (b) SM3: 3840 RU, (c) SM4s: 3720 RU, (d) SM4s: 3620 RU, (e) PC: 3550 RU, (f) PC: 3530 RU, (g) SM2a: 3450 RU, (h) SM2a: 3390 RU, (i) GM2: 3340 RU, (j) GM2: 3320 RU, (k) SB1a: 3260 RU, (l) SB1a: 3240 RU, (m) SB2: 3210 RU, (n) SB2: 3170 RU, (o) SM3: 3130 RU. (A) Linear scale; (B) logarithmic scale of same date with base line set to 50 RU. (For interpretation of the reference to colour in this figure legend, the reader is referred to the web version of this article.)
58
R. Sandhoff et al. / Biochimica et Biophysica Acta 1687 (2005) 52–63
Fig. 5. Relative binding of sulfatide-liposomes to wild-type MCP-1/CCL2 and 18AA19-MCP-1/CCL2. The chemokines were immobilized onto a CM5-sensor chip surface (MCP-1/CCL2: 1860 RU, and 18AA19-MCP-1/ CCL2: 990 and 2760 RU). Liposomes containing 20 mol% cholesterol, 55 mol% PC and 25 mol% SM4s, SM3, SB2, or SB1a were injected for 150 s in a liposomal lipid concentration of 150 AM and a flow rate of 20 Al/min. Running and injection buffer 50 mM Tris, 150 mM NaCl, pH 7.4. Binding values were taken at 145-s injection time and plotted relative to the corresponding values of sulfatide–wild-type MCP-1/CCL2 interaction. Bars represent the mean of two measurements. The data range is depicted by the lines on top of the bars. The mean of the absolute value of binding to MCP1/CCL2 was set to 100%.
chemokines. Whereas MCP-1/CCL2 bound to all sulfatides tested (SM4s, SM3, SM2a, SB2 and SB1a), IL-8/CXCL8 bound SM3, SM2a, and SB2; SDF-1a/CXCL12 bound SM3, SM2a, SB2, and SB1a; MIP-1h/CCL4 was selective to SM3, SM2a and SB1a; and MIP-1a/CCL3 to SM4s, SM3, SM2a, SB2, and SB1a. Neither PC- nor LacCerliposomes interacted significantly with the chemokines tested and ganglioside binding was low (Fig. 6-I). A comparison of all the chemokines immobilized at an amount of 1200 to 1700 RU showed sulfatide SM3 always to have the highest binding capacity irrespective of its molecular sulfate content, followed by SM2a, SB1a and SB2 (Fig. 6II) (for MIP-1a/CCL3 only data with V300 RU immobilization rate were obtained). Interestingly, among the gangliosides, GM3, the sialo-analog of SM3, showed the highest affinity to chemokines, which was in the range of SM4s binding. Nevertheless, comparing GSL of the same neutral oligosaccharide core structure, the sulfatides displayed at least 2.5-fold higher affinities to each chemokine than their corresponding sialic-acid derivatives. The structurally related neutral glycosphingolipids revealed even less binding to the chemokines (a: SM3NGM3NLacCer, b: SM2aNGM2) (Fig. 6-II).
4. Discussion Sulfatides are anionic and amphiphilic glycosphingolipids (GSLs) found in the extracellular lipid bilayer leaflet of the cell plasma membrane that are thought to participate in events of cell–cell and cell–matrix contacts. GSLs are enriched in subcellular fractions obtained in the presence of detergents that are interpreted as remnants of lipid-rafts that
appear to function as signaling platforms for cells [36]. In this context, GSLs may be able to affect receptor–ligand interactions by either modulating the affinity of a receptor for its ligand or by binding its ligand [37]. GSLs are regulated during cell cycle, cell differentiation and activation and have been described to modulate signal transduction through their interaction with receptors [38–51]. Sulfatides are up-regulated in cells exposed to osmotic stress [52]. Transcripts of the cerebroside-sulfotransferase (CST), responsible for GSL sulfation, are tissue-specifically spliced and CST-protein activity may be regulated posttranslationally [53]. Sulfatides have been described to influence the production of cytokines in human mononuclear cells [54]. In addition, they appear to negatively regulate the expression of myelin basic protein [55] and the expression of integrins [50]. Sulfatides have been reported to bind specifically to cell adhesive proteins including thrombospondins, laminins [56] and selectins [57], as well as regulator proteins of the alternative pathway of complement activation [20]. The sulfatide binding capacity of these proteins has been linked to the presence of basic amino acid clusters (XBBXBX, BXBXBX, B represents a basic residue) or the consensus sequence CSVTCGXGXXXRXR that is responsible not only for SM4s but also for heparin binding [52]. Basic amino acids or clusters are also important for GAG binding of chemokines. The GAG-binding regions of the inflammatory chemokines MIP-1a/CCL3, MIP-1h/CCL4, and RANTES/CCL5 as well as the homeostatic chemokine SDF-1a/CXCL12 involve a BBXB-cluster [9,10,58,59]. In IL-8/CXCL8, K64 and R68 of the C-terminal alpha-helix and K20 in the loop around residues 18–23 [60] contribute to GAG binding, whereas in MCP-1/CCL2 the basic amino acids R18, K19, R24, K49, K58 H66 are involved [3,7,61]. These residues could also potentially contribute to sulfatide binding of chemokines. This report demonstrates that the chemokines MCP-1/ CCL2, IL-8/CXCL8, SDF-1a/CXCL12, and MIP-1h/CCL4 bind to sulfatide containing liposomes. MIP-1a/CCL3 showed comparably lower but nevertheless specific binding to sulfatides. The role of basic residues, important for GAG binding, were shown to contribute to the sulfatide binding as was demonstrated with the R18A,K19A mutant of MCP-1/ CCL2 for which the sulfatide interaction was indeed significantly diminished. Importantly, for MCP-1/CCL2, coupling to the chip surface via amino groups or via the histag gave similar results, demonstrating no essential influence of the coupling method in this assay. Liposomes containing the complex sulfatides SM3, SM2a, SB2, and SB1a displayed the highest binding to the chemokines, whereas SM4s containing liposomes bound only to MCP-1/ CCL2 in a reproducible manner. In general, SM3 showed the highest affinity for chemokines. Although SM4s, SM3 and SB1a carry the same terminal galactosyl-3-sulfate epitope, the low binding capacity of SM4s for chemokines is conspicuous. This discrepancy in binding capacity
R. Sandhoff et al. / Biochimica et Biophysica Acta 1687 (2005) 52–63
probably originates from differences in the way this epitope is presented. In SM2a as well as SB2, the participation of the sulfate group may be altered by an intramolecular hydrogen bonding with the neighboring GalNAc residue. The relative independence of the chemokine-sulfatide complexation from pH as well as the presence of divalent cations suggests that factors in addition to electric charge interactions may play a role. This is corroborated by the finding that the MCP-1/CCL2-basic amino acid mutant retains part of its capacity to bind sulfatides, as well as the fact that the various sulfatides show chemokine complexation capabilities that do not solely correspond to their sulfate content. Nevertheless, essentially no chemokine binding was observed with liposomes containing either the structurally related acidic sialo-GSL, neutral GSL, or the
59
neutral or acidic phospholipids PS and PI, emphasizing the role of sulfate groups. A differential dependency of chemokine-sulfatide interactions on chemokine concentration, as observed for MCP-1/CCL2 with a more logarithmic and SDF-1a/CXCL12 with a more linear dependency, may reflect the necessity of some (e.g., MCP-1/CCL2) but not all (e.g., SDF-1a/CXCL12) chemokines to form oligomers in order to bind sulfatides with high affinity. Sulfatides are synthesized in the lumen of the Golgiapparatus and transported to the cell surface. Hence, the sulfated oligosaccharide head-group of these compounds is exposed to the extracellular side of the plasma membrane, a prerequisite for a possible interaction with extracellular chemokines. It appears likely that thereby sulfatides may influence/modulate the outcome of chemokine effects on
Fig. 6. SPR sensogram: sulfatides containing liposomes specifically bind to chemokines MCP-1/CCL2, IL-8/CXCL8, SDF-1a/CXCL12, MIP-1h/CCL4 and MIP-1a/CCL3. (I) The chemokines were immobilized onto a CM5-sensor chip surface; (A) IL-8/CXCL8: 2400 RU, (B) MCP-1/CCL2: 2790 RU, (C) SDF-1a/ CXCL12: 790 RU, (D) IL-8/CXCL8: 1700 RU, (E) MCP-1/CCL2: 1550 RU, (F) SDF-1a/CXCL12: 1200 RU, (G) MIP-1h/CCL4: 1380 RU, (H) MIP-1h/ CCL4: 400 RU, (I) MIP-1a/CCL3: 300 RU. Liposomes containing 20 mol% cholesterol, 55 mol% PC and 25 mol% of the indicated lipid were injected (start: 0 s) for 150 s in a liposomal lipid concentration of 150 AM and at a flow rate of 20 Al/min [indicated lipid: PC (light green), LacCer (light cyan), GM3 (dark yellow), GM2 (gray), GM1 (light gray), GD1a (yellow), SM4s (dark red), SM3 (dark blue), SM2a (light magenta), SB2 (light red) and SB1a (magenta)]. Running and injection buffer: 50 mM Tris, 150 mM NaCl, pH 7.4. Spikes at 0 and 150 s are due to the pumps and IFC system. (II) Diagram comparing the amount of liposomes bound after 145 s of injection time to chemokines immobilized at a rate of 1200 to 1700 RU (for MIP-1a/CCL3 only data with V300 RU immobilization rate were obtained). Data were obtained from figures 6-I D, E, F, G and I. Bars represent the mean of two measurements. The data range is depicted by the lines on top of the bars. The mean of the absolute value of SM3 binding was set to 100%.
60
R. Sandhoff et al. / Biochimica et Biophysica Acta 1687 (2005) 52–63
Fig. 6 (continued).
chemotaxis, cell proliferation and angiogenesis [62–66]. Sulfatides, like other glycosphingolipids (GSL), are believed to concentrate in raft-microdomains of the plasma membrane. The chemokine-receptors CCR5, CXCR4 and CCR2 have also been described to be within rafts [1]. Modulation of plasma membrane events by GSL has been shown to play a role in the internalization of HIV-1 gp120 after binding to CD4 and CXCR4/CCR5 [67]. Thus, binding of chemokines to sulfatides in the proximity of receptors could result in increasing local chemokine concentration and a prolonged intracellular signaling of cells expressing sulfatides (cis-action). Alternatively direct cell–cell interactions through corresponding sphingolipid enriched signaling platforms could be facilitated (transaction). Human peripheral blood neutrophils produce and react to IL-8/CXCL8 while they shed most of their sulfatides to the outside medium [22]. It is intriguing to speculate that this shedding of sulfatides may serve similar functions as it seems the case for soluble cytokine-receptors. Their extracellular cytokine-binding domains can be cleaved off by proteases, then acting as down-regulators of cytokinesignaling making them specific therapeutic agents in human disease. Alternatively, soluble cytokine-receptors act as carriers thereby stabilizing the chemokine, giving them a prolonged half life, which then might result in potentiation of the cytokine activity [68,69]. GSLs, normally integral molecules of the plasma membrane, are shed extensively into the extracellular surrounding by tumor cells [70,71]. In renal carcinoma cells, elevated concentrations of sulfatides including the more complex structures SM3, SM2, and SB2 are found
[26,27,72]. Chemokines are known to influence tumor biology in many ways, e.g., tumor growth, angiogenesis or monocyte/macrophage infiltration [73–75]. In addition, tumor cells are also known to produce chemokines by themselves [76]. Binding of sulfatides to chemokines may already occur within the Golgi or trans-Golgi network of these tumor cells. This could, perhaps, result in secretion as well as presentation of sulfatide-bound chemokines, leading to a modulation of tumor-growth. In summary we could demonstrate that, besides binding to GAGs or receptor proteins, chemokines are able to bind specifically to sulfatides, sulfated glycosphingolipids at the plasma membrane, thereby possibly enhancing the presentation of chemokines.
Acknowledgement We thank Benita von Tqmpling-Radosta for her engagement in this study and Dr. Thomas Bqch for his helpful comments during the proceedings of this work. The work was supported by a grant from the DFG: SFB 405, B10 to H.-J.G. and SFB 469, B2 and SFB 571, C2 to P.J.N.
References [1] S. Manes, R.A. Lacalle, C. Gomez-Mouton, G. del Real, E. Mira, A.C. Martinez, Membrane raft microdomains in chemokine receptor function, Semin. Immunol. 13 (2001) 147 – 157.
R. Sandhoff et al. / Biochimica et Biophysica Acta 1687 (2005) 52–63 [2] S. Ali, A.C. Palmer, B. Banerjee, S.J. Fritchley, J.A. Kirby, Examination of the function of RANTES, MIP-1alpha, and MIP1beta following interaction with heparin-like glycosaminoglycans, J. Biol. Chem. 275 (2000) 11721 – 11727. [3] A.E. Proudfoot, T.M. Handel, Z. Johnson, E.K. Lau, P. LiWang, I. Clark-Lewis, F. Borlat, T.N. Wells, M.H. Kosco-Vilbois, Glycosaminoglycan binding and oligomerization are essential for the in vivo activity of certain chemokines, Proc. Natl. Acad. Sci. U. S. A. 100 (2003) 1885 – 1890. [4] G.S. Kuschert, F. Coulin, C.A. Power, A.E. Proudfoot, R.E. Hubbard, A.J. Hoogewerf, T.N. Wells, Glycosaminoglycans interact selectively with chemokines and modulate receptor binding and cellular responses, Biochemistry 38 (1999) 12959 – 12968. [5] M. Baggiolini, P. Loetscher, Chemokines in inflammation and immunity, Immunol. Today 21 (2000) 418 – 420. [6] A.J. Hoogewerf, G.S. Kuschert, A.E. Proudfoot, F. Borlat, I. ClarkLewis, C.A. Power, T.N. Wells, Glycosaminoglycans mediate cell surface oligomerization of chemokines, Biochemistry 36 (1997) 13570 – 13578. [7] L. Chakravarty, L. Rogers, T. Quach, S. Breckenridge, P.E. Kolattukudy, Lysine 58 and histidine 66 at the C-terminal alpha-helix of monocyte chemoattractant protein-1 are essential for glycosaminoglycan binding, J. Biol. Chem. 273 (1998) 29641 – 29647. [8] J.M. Burns, R.C. Gallo, A.L. DeVico, G.K. Lewis, A new monoclonal antibody, mAb 4A12, identifies a role for the glycosaminoglycan (GAG) binding domain of RANTES in the antiviral effect against HIV-1 and intracellular Ca2+ signaling, J. Exp. Med. 188 (1998) 1917 – 1927. [9] A. Amara, O. Lorthioir, A. Valenzuela, A. Magerus, M. Thelen, M. Montes, J.L. Virelizier, M. Delepierre, F. Baleux, H. Lortat-Jacob, F. Arenzana-Seisdedos, Stromal cell-derived factor-1alpha associates with heparan sulfates through the first beta-strand of the chemokine, J. Biol. Chem. 274 (1999) 23916 – 23925. [10] A.E. Proudfoot, S. Fritchley, F. Borlat, J.P. Shaw, F. Vilbois, C. Zwahlen, A. Trkola, D. Marchant, P.R. Clapham, T.N. Wells, The BBXB motif of RANTES is the principal site for heparin binding and controls receptor selectivity, J. Biol. Chem. 276 (2001) 10620 – 10626. [11] L. Martin, C. Blanpain, P. Garnier, V. Wittamer, M. Parmentier, C. Vita, Structural and functional analysis of the RANTES-glycosaminoglycans interactions, Biochemistry 40 (2001) 6303 – 6318. [12] W. Koopmann, M.S. Krangel, Identification of a glycosaminoglycanbinding site in chemokine macrophage inflammatory protein-1alpha, J. Biol. Chem. 272 (1997) 10103 – 10109. [13] W. Koopmann, C. Ediriwickrema, M.S. Krangel, Structure and function of the glycosaminoglycan binding site of chemokine macrophage-inflammatory protein-1 beta, J. Immunol. 163 (1999) 2120 – 2127. [14] T. Kobayashi, K. Honke, T. Miyazaki, K. Matsumoto, T. Nakamura, I. Ishizuka, A. Makita, Hepatocyte growth factor specifically binds to sulfoglycolipids, J. Biol. Chem. 269 (1994) 9817 – 9821. [15] D.D. Roberts, Interactions of thrombospondin with sulfated glycolipids and proteoglycans of human melanoma cells, Cancer Res. 48 (1988) 6785 – 6793. [16] T. Kobayashi, K. Honke, Y. Kuramitsu, M. Hosokawa, T. Miyazaki, J. Murata, I. Saiki, I. Ishizuka, A. Makita, Cell-surface sulfoglycolipids are involved in the attachment of renal-cancer cells to laminin, Int. J. Cancer 56 (1994) 281 – 285. [17] R. Miura, A. Aspberg, I.M. Ethell, K. Hagihara, R.L. Schnaar, E. Ruoslahti, Y. Yamaguchi, The proteoglycan lectin domain binds sulfated cell surface glycolipids and promotes cell adhesion, J. Biol. Chem. 274 (1999) 11431 – 11438. [18] M.S. Mulligan, R.L. Warner, J.B. Lowe, P.L. Smith, Y. Suzuki, M. Miyasaka, S. Yamaguchi, Y. Ohta, Y. Tsukada, M. Kiso, A. Hasegawa, P.A. Ward, In vitro and in vivo selectin-blocking activities of sulfated lipids and sulfated sialyl compounds, Int. Immunol. 10 (1998) 569 – 575.
61
[19] J. Kajihara, Y. Guoji, K. Kato, Y. Suzuki, Sulfatide, a specific sugar ligand for L-selectin, blocks CCl4-induced liver inflammation in rats, Biosci. Biotechnol. Biochem. 59 (1995) 155 – 157. [20] R. Jennemann, M. Schulze, B.L. Bauer, C. Kurtz, H. Wiegandt, Serum immunoglobulins in Heymann’s experimental nephritis modulate binding of properdin and factor-H to sulfo-glycosphingolipids II3SO3( )-Gg3Cer and III3SO3( )-, II3SO3( )–Gg3Cer, J. Biochem. (Tokyo) 116 (1994) 450 – 456. [21] L.L. Sarlieve, B. Zalc, N.M. Neskovic, J.P. Zanetta, G. Rebel, Structure and immunological localization of spleen sulfolipid, Biochim. Biophys. Acta 795 (1984) 166 – 168. [22] A. Aruffo, W. Kolanus, G. Walz, P. Fredman, B. Seed, CD62/Pselection recognition of myeloid and tumor cell sulfatides, Cell 67 (1991) 35 – 44. [23] C.G. Hansson, K.A. Karlsson, B.E. Samuelsson, The identification of sulfatides in human erythrocyte membrane and their relation to sodium-potassium dependent adenosine triphosphatase, J. Biochem. (Tokyo) 83 (1978) 813 – 819. [24] Y. Kushi, M. Arita, I. Ishizuka, T. Kasama, P. Fredman, S. Handa, Sulfatide is expressed in both erythrocytes and platelets of bovine origin, Biochim. Biophys. Acta 1304 (1996) 254 – 262. [25] N. Hiraiwa, Y. Fukuda, H. Imura, K. Tadano-Aritomi, K. Nagai, I. Ishizuka, R. Kannagi, Accumulation of highly acidic sulfated glycosphingolipids in human hepatocellular carcinoma defined by a series of monoclonal antibodies, Cancer Res. 50 (1990) 2917 – 2928. [26] T. Kobayashi, Sulfolipids and glycolipid sulfotransferases in a human renal cell carcinoma cell line, Hokkaido Igaku Zasshi 67 (1992) 522 – 530. [27] T. Kobayashi, K. Honke, K. Kamio, N. Sakakibara, S. Gasa, N. Miyao, T. Tsukamoto, I. Ishizuka, T. Miyazaki, A. Makita, Sulfolipids and glycolipid sulfotransferase activities in human renal cell carcinoma cells, Br. J. Cancer 67 (1993) 76 – 80. [28] D. Mamelak, M. Mylvaganam, H. Whetstone, E. Hartmann, W. Lennarz, P.B. Wyrick, J. Raulston, H. Han, P. Hoffman, C.A. Lingwood, Hsp70s contain a specific sulfogalactolipid binding site. Differential aglycone influence on sulfogalactosyl ceramide binding by recombinant prokaryotic and eukaryotic hsp70 family members, Biochemistry 40 (2001) 3572 – 3582. [29] H. Ideo, A. Seko, I. Ishizuka, K. Yamashita, The N-terminal carbohydrate recognition domain of galectin-8 recognizes specific glycosphingolipids with high affinity, Glycobiology 13 (2003) 713 – 723. [30] R. Sandhoff, S.T. Hepbildikler, R. Jennemann, R. Geyer, V. Gieselmann, R.L. Proia, H. Wiegandt, H.J. Grone, Kidney sulfatides in mouse models of inherited glycosphingolipid disorders. Determination by nano-electrospray ionization tandem mass spectrometry, J. Biol. Chem. 277 (2002) 20386 – 20398. [31] R.G. Spiro, Analysis of sugars found in glycoproteins; determination of neutral sugars; total hexoses by anthrone reaction, Methods Enzymol. VIII (1966) 4 – 5. [32] G.M. Kuziemko, M. Stroh, R.C. Stevens, Cholera toxin binding affinity and specificity for gangliosides determined by surface plasmon resonance, Biochemistry 35 (1996) 6375 – 6384. [33] C.R. MacKenzie, T. Hirama, K.K. Lee, E. Altman, N.M. Young, Quantitative analysis of bacterial toxin affinity and specificity for glycolipid receptors by surface plasmon resonance, J. Biol. Chem. 272 (1997) 5533 – 5538. [34] M.A. Cooper, A. Hansson, S. Lofas, D.H. Williams, A vesicle capture sensor chip for kinetic analysis of interactions with membrane-bound receptors, Anal. Biochem. 277 (2000) 196 – 205. [35] L. Nieba, S.E. Nieba-Axmann, A. Persson, M. Hamalainen, F. Edebratt, A. Hansson, J. Lidholm, K. Magnusson, A.F. Karlsson, A. Pluckthun, BIACORE analysis of histidine-tagged proteins using a chelating NTA sensor chip, Anal. Biochem. 252 (1997) 217 – 228.
62
R. Sandhoff et al. / Biochimica et Biophysica Acta 1687 (2005) 52–63
[36] K. Simons, E. Ikonen, Functional rafts in cell membranes, Nature 387 (1997) 569 – 572. [37] S.I. Hakomori Si, Inaugural Article: the glycosynapse, Proc. Natl. Acad. Sci. U. S. A. 99 (2002) 225 – 232. [38] I. Majoul, T. Schmidt, M. Pomasanova, E. Boutkevich, Y. Kozlov, H.D. Soling, Differential expression of receptors for Shiga and Cholera toxin is regulated by the cell cycle, J. Cell. Sci. 115 (2002) 817 – 826. [39] H. Nojiri, F. Takaku, Y. Terui, Y. Miura, M. Saito, Ganglioside GM3: an acidic membrane component that increases during macrophage-like cell differentiation can induce monocytic differentiation of human myeloid and monocytoid leukemic cell lines HL-60 and U937, Proc. Natl. Acad. Sci. U. S. A. 83 (1986) 782 – 786. [40] B.K. Gillard, M.A. Jones, A.A. Turner, D.E. Lewis, D.M. Marcus, Interferon-gamma alters expression of endothelial cell-surface glycosphingolipids, Arch. Biochem. Biophys. 279 (1990) 122 – 129. [41] Z. Ding, H. Kawashima, Y. Suzuki, T. Suzuki, P.A. Ward, M. Miyasaka, A sulfatide receptor distinct from L-selectin is involved in lymphocyte activation, FEBS Lett. 418 (1997) 310 – 314. [42] Z.Q. Ding, H. Kawashima, M. Miyasaka, Sulfatide binding and activation of leukocytes through an L-selectin-independent pathway, J. Leukoc. Biol. 68 (2000) 65 – 72. [43] G.F. Sud’ina, T.G. Brock, M.A. Pushkareva, S.I. Galkina, D.V. Turutin, M. Peters-Golden, V. Ullrich, Sulfatides trigger polymorphonuclear granulocyte spreading on collagen-coated surfaces and inhibit subsequent activation of 5-lipoxygenase, Biochem. J. 359 (2001) 621 – 629. [44] A. Karlsson, H. Miller-Podraza, P. Johansson, K.A. Karlsson, C. Dahlgren, S. Teneberg, Different glycosphingolipid composition in human neutrophil subcellular compartments, Glycoconj. J. 18 (2001) 231 – 243. [45] J.F. Bukowski, M.G. Roncarolo, H. Spits, M.S. Krangel, C.T. Morita, M.B. Brenner, H. Band, T cell receptor-dependent activation of human lymphocytes through cell surface ganglioside GT1b: implications for innate immunity [In Process Citation], Eur. J. Immunol. 30 (2000) 3199 – 3206. [46] K. Iwabuchi, I. Nagaoka, Lactosylceramide-enriched glycosphingolipid signaling domain mediates superoxide generation from human neutrophils, Blood 100 (2002) 1454 – 1464. [47] J.I. Inokuchi, S. Uemura, K. Kabayama, Y. Igarashi, Glycosphingolipid deficiency affects functional microdomain formation in Lewis lung carcinoma cells, Glycoconj. J. 17 (2000) 239 – 245. [48] E.J. Meuillet, B. Mania-Farnell, D. George, J.I. Inokuchi, E.G. Bremer, Modulation of EGF receptor activity by changes in the GM3 content in a human epidermoid carcinoma cell line, A431, Exp. Cell Res. 256 (2000) 74 – 82. [49] S. Tagami, J. Inokuchi Ji, K. Kabayama, H. Yoshimura, F. Kitamura, S. Uemura, C. Ogawa, A. Ishii, M. Saito, Y. Ohtsuka, S. Sakaue, Y. Igarashi, Ganglioside GM3 participates in the pathological conditions of insulin resistance, J. Biol. Chem. 277 (2002) 3085 – 3092. [50] K. Kabayama, N. Ito, K. Honke, Y. Igarashi, J. Inokuchi, Suppression of integrin expression and tumorigenicity by sulfation of lactosylceramide in 3LL Lewis lung carcinoma cells, J. Biol. Chem. 276 (2001) 26777 – 26783. [51] S. Uemura, K. Kabayama, M. Noguchi, Y. Igarashi, J. Inokuchi, Sialylation and sulfation of lactosylceramide distinctly regulate anchorage-independent growth, apoptosis, and gene expression in3LL Lewis lung carcinoma cells, Glycobiology 13 (2003) 207 – 216. [52] I. Ishizuka, Chemistry and functional distribution of sulfoglycolipids, Prog. Lipid Res. 36 (1997) 245 – 319. [53] Y. Hirahara, M. Tsuda, Y. Wada, K. Honke, cDNA cloning, genomic cloning, and tissue-specific regulation of mouse cerebroside sulfotransferase, Eur. J. Biochem. 267 (2000) 1909 – 1917. [54] K. Buschard, M. Diamant, L.E. Bovin, J.E. Mansson, P. Fredman, K. Bendtzen, Sulfatide and its precursor galactosylceramide influence the production of cytokines in human mononuclear cells, APMIS 104 (1996) 938 – 944.
[55] R. Bansal, S. Winkler, S. Bheddah, Negative regulation of oligodendrocyte differentiation by galactosphingolipids, J. Neurosci. 19 (1999) 7913 – 7924. [56] D.D. Roberts, Sulfatide-binding proteins, Methods Enzymol. 138 (1987) 473 – 483. [57] A. Varki, Selectin ligands, Proc. Natl. Acad. Sci. U. S. A. 91 (1994) 7390 – 7397. [58] G.J. Graham, P.C. Wilkinson, R.J. Nibbs, S. Lowe, S.O. Kolset, A. Parker, M.G. Freshney, M.L. Tsang, I.B. Pragnell, Uncoupling of stem cell inhibition from monocyte chemoattraction in MIP-1alpha by mutagenesis of the proteoglycan binding site, EMBO J. 15 (1996) 6506 – 6515. [59] J.S. Laurence, C. Blanpain, A. De Leener, M. Parmentier, P.J. LiWang, Importance of basic residues and quaternary structure in the function of MIP-1 beta: CCR5 binding and cell surface sugar interactions, Biochemistry 40 (2001) 4990 – 4999. [60] G.S. Kuschert, A.J. Hoogewerf, A.E. Proudfoot, C.W. Chung, R.M. Cooke, R.E. Hubbard, T.N. Wells, P.N. Sanderson, Identification of a glycosaminoglycan binding surface on human interleukin-8, Biochemistry 37 (1998) 11193 – 11201. [61] E.K. Lau, C.D. Paavola, Z. Johnson, J.P. Gaudry, E. Geretti, F. Borlat, A.J. Kungl, A.E. Proudfoot, T.M. Handel, Identification of the glycosaminoglycan binding site of the CC chemokine, MCP-1: implications for structure and function in vivo, J. Biol. Chem. 279 (2004) 22294 – 22305. [62] A. Li, S. Dubey, M.L. Varney, B.J. Dave, R.K. Singh, IL-8 directly enhanced endothelial cell survival, proliferation, and matrix metalloproteinases production and regulated angiogenesis, J. Immunol. 170 (2003) 3369 – 3376. [63] H.E. Broxmeyer, L. Kohli, C.H. Kim, Y. Lee, C. Mantel, S. Cooper, G. Hangoc, M. Shaheen, X. Li, D.W. Clapp, Stromal cell-derived factor-1/CXCL12 directly enhances survival/antiapoptosis of myeloid progenitor cells through CXCR4 and G(alpha)i proteins and enhances engraftment of competitive, repopulating stem cells, J. Leukoc. Biol. 73 (2003) 630 – 638. [64] R. Pakala, T. Watanabe, C.R. Benedict, Induction of endothelial cell proliferation by angiogenic factors released by activated monocytes, Cardiovasc. Radiat. Med. 3 (2002) 95 – 101. [65] B. Banas, M. Wornle, T. Berger, P.J. Nelson, C.D. Cohen, M. Kretzler, J. Pfirstinger, M. Mack, M. Lipp, H.J. Grone, D. Schlondorff, Roles of SLC/CCL21 and CCR7 in human kidney for mesangial proliferation, migration, apoptosis, and tissue homeostasis, J. Immunol. 168 (2002) 4301 – 4307. [66] J.A. Belperio, M.P. Keane, D.A. Arenberg, C.L. Addison, J.E. Ehlert, M.D. Burdick, R.M. Strieter, CXC chemokines in angiogenesis, J. Leukoc. Biol. 68 (2000) 1 – 8. [67] P. Hug, H.M. Lin, T. Korte, X. Xiao, D.S. Dimitrov, J.M. Wang, A. Puri, R. Blumenthal, Glycosphingolipids promote entry of a broad range of human immunodeficiency virus type 1 isolates into cell lines expressing CD4, CXCR4, and/or CCR5, J. Virol. 74 (2000) 6377 – 6385. [68] R. Fernandez-Botran, Soluble cytokine receptors: basic immunology and clinical applications, Crit. Rev. Clin. Lab. Sci. 36 (1999) 165 – 224. [69] M.L. Heaney, D.W. Golde, Soluble receptors in human disease, J. Leukoc. Biol. 64 (1998) 135 – 146. [70] S. Ladisch, B. Gillard, C. Wong, L. Ulsh, Shedding and immunoregulatory activity of YAC-1 lymphoma cell gangliosides, Cancer Res. 43 (1983) 3808 – 3813. [71] S. Ladisch, S. Kitada, E.F. Hays, Gangliosides shed by tumor cells enhance tumor formation in mice, J. Clin. Invest. 79 (1987) 1879 – 1882. [72] T. Kobayashi, K. Honke, S. Gasa, Y. Saga, T. Miyazaki, K. TadanoAritomi, I. Ishizuka, A. Makita, Enzymatic sulfation of gangliotriaosylceramide in human renal cancer cells, J. Biochem. (Tokyo) 117 (1995) 987 – 992.
R. Sandhoff et al. / Biochimica et Biophysica Acta 1687 (2005) 52–63 [73] A.P. Vicari, C. Caux, Chemokines in cancer, Cytokine Growth Factor Rev. 13 (2002) 143 – 154. [74] D.T. Graves, A.J. Valente, Monocyte chemotactic proteins from human tumor cells, Biochem. Pharmacol. 41 (1991) 333 – 337. [75] M. Platten, A. Kretz, U. Naumann, S. Aulwurm, K. Egashira, S. Isenmann, M. Weller, Monocyte chemoattractant protein-1 increases
63
microglial infiltration and aggressiveness of gliomas, Ann. Neurol. 54 (2003) 388 – 392. [76] R. Salcedo, J.J. Oppenheim, Role of chemokines in angiogenesis: CXCL12/SDF-1 and CXCR4 interaction, a key regulator of endothelial cell responses, Microcirculation 10 (2003) 359 – 370.