Introduction

Trafficking of hematopoietic progenitor cells (HPC) towards the bone marrow after clinical transplantation procedures requires the specific interaction between adhesion molecules and receptors on HPC with their ligands on bone marrow endothelial cells (BMEC), the gatekeepers between the bone marrow stroma and the sinusoidal lumen.^1,2 However, the precise mechanisms of this homing process are still unclear. It is thought that comparable with leukocyte extravasation in inflammation or lymphocyte homing to lymph nodes, a multistep process occurs,^3,4,5 in which selectin-mediated rolling, integrin activation and firm adhesion followed by transendothelial migration are the main successive steps. Regarding the interaction between HPC and BMEC it has been shown that in vivo rolling is mediated by endothelial E-selectin, P-selectin and vascular cell adhesion molecule-1 (VCAM-1).^6,7 These adhesion molecules are constitutively expressed on BMEC,⁸ whereas expression of these molecules on endothelium elsewhere in the body is associated with an activated phenotype. It has been described by a number of groups that the firm adhesion step of HPC to BMEC or other endothelial cells is mediated via binding of the integrins very late activation antigen-4 (VLA-4), VLA-5 and leukocyte function-associated antigen-1 (LFA-1) on the HPC to their endothelial receptors, respectively VCAM-1 and intercellular adhesion molecule-1 (ICAM-1).^9,10,11 Recently, Peled et al^9,12 reported that the integrins VLA-4 and LFA-1 on umbilical cord blood CD34⁺ cells can be activated via signaling of the CXCR4 receptor when the chemokine stromal cell-derived factor-1 (SDF-1) is present on endothelial cells or integrin-binding substrates, resulting in firm adhesion and spreading of the cells.

SDF-1 is a CXC chemokine which is present in many tissues.^13,14,15 It has been reported that SDF-1 can act as a major chemoattractant for HPC and SCID repopulating cells^16,17 as well as leukemic cells,^18,19 via its receptor CXCR4,^13,20 which can signal via G proteins of the G_i class resulting in migration of the cells. It has been shown by using gene targeting experiments in mice that SDF-1 and CXCR4-deficiencies are lethal and lead to a defect in hematopoiesis.^21,22 Blocking CXCR4 of SCID repopulating cells transplanted in NOD/SCID mice prevented engraftment.^17,23 These results all suggest that SDF-1 and CXCR4 play a critical role in HPC homing. In the bone marrow, SDF-1 is produced by stromal cells and osteoblasts.^13,24 It has been shown by immunohistochemistry that SDF-1 is present on the luminal side of BMEC.²⁵ The identity of the SDF-1 binding site on BMEC however, is not known. SDF-1 has been reported by other groups to be able to bind to cellular proteoglycans (PGs).^26,27 PGs are macromolecules consisting of a core protein covalently attached to linear glycosaminoglycan (GAG) chains. These polysaccharide GAG chains consist of repeating disaccharide units. The composition of repeating disaccharides determines the type of GAGs. Heparan sulfate (HS)/heparin-like GAGs consist of the combination of glucosamine with an uronic acid (either glucuronic acid or its epimer iduronic acid), whereas chondroitin and dermatan sulfate (CS/DS) GAGs consist of galactosamine with a glucuronic or iduronic acid residue respectively. The HS-polysaccharide chain is substituted to a varying extent with N- and O-linked sulfate groups. The sulfation pattern of a GAG chain determines the ability to bind different proteins. We reported recently that cell-associated HS-GAGs from the BMEC cell line 4LHBMEC contained higher sulfated domains compared to similar GAGs from human umbilical vein endothelial cells (HUVEC).²⁸ Furthermore, it was found that 4LHBMEC bound more ¹²⁵I-SDF-1 compared to HUVEC per cell. This binding could be inhibited by competition with heparin, O-sulfated heparin (N-desulfated, N-acetylated heparin) and to a lesser extent by N-sulfated heparin (completely desulfated, N-sulfated heparin), suggesting a role for an O-sulfated motif in endothelial HS-GAGs in the binding of SDF-1. The aim of the present study was to further investigate the binding of SDF-1 to BMEC via HS-GAGs and/or other GAGs. Furthermore, it was studied whether PGs on BMEC could play a role in firm arrest of HPC under flow conditions via binding and presentation of SDF-1 and could thus play a key role in the homing process of HPC.

Materials and methods

GAGs and antibodies

GAGs used were heparin (Bufa Chemie, Uitgeest, The Netherlands), low molecular weight heparin: dalteparin (Fragmin, Pharmacia, Peapack, NJ, USA), HS from bovine intestinal mucosa (HSIM), HS sulfate from bovine kidney (HSBK), dermatan sulfate (DS), chondroitin sulfate A (CS) (all from Sigma, St Louis, MO, USA). Radioactive labeled GAGs used were ³H-heparin and ³H-HSBK (prepared according to Hööket al²⁹ using ³H-acetic anhydride). Antibodies used were: anti-CXCR4-PE clone 12G5 (1:5) (PharMingen, San Diego, Ca, USA), anti-SDF-1 MAB310, clone 79014.111 (R&D Systems, Abingdon, UK) and anti-SDF-1 K15C, kindly provided by F Arenzana-Seisdedos (Paris, France).²⁶ As second step antibody rabbit anti-mouse-Ig-PE (Dako, Glostrup, Denmark) was used.

Cell cultures

Human BMEC cell lines 4LHBMEC²⁸ (used in most experiments), TrHBMEC¹⁰ and HBMEC60³⁰ (kindly provided by Dr CE van der Schoot, CLB, Amsterdam, The Netherlands) were cultured in endothelial cell culture medium consisting of medium 199, 10% new born calf serum, 1% glutamine (Gibco, Grand Island, NY, USA), 10% human pooled serum (CLB, or BioWhittaker, Walkersville, MD, USA), 0.15 mg/ml crude endothelial cell growth factor and 5 IU/ml heparin at 37°C and 5% CO₂ in six-well plates coated with fibronectin (kindly provided by Dr JA van Mourik (CLB)) and passaged twice a week by detachment by trypsin-EDTA (BioWhittaker). HUVEC were isolated as described by Jaffe et al³¹ from umbilical cords obtained from the department of Gynecology and Obstetrics (Vrije Universiteit Medical Center, Amsterdam, The Netherlands) after informed consent. HUVEC were used until passage four. Three HPC cell lines were used: the myeloblastic CD34⁺, CXCR4⁺ cell line KG-1 (American type culture collection (ATCC) CCL-246, Rockville, MD, USA), cultured in IMDM with 20% fetal calf serum, 1% L-glutamine and 50 M -mercapto-ethanol (all from Gibco), the CD34⁺, CXCR4^- cell line KG-1a (ATCC CCL-246.1) and KG-1v, a CD34^-, high CXCR4-expressing variant of the KG-1 cell line.³² KG-1a and KG-1v were maintained in RPMI-1640 (Gibco), supplemented with 10% fetal calf serum. Human bone marrow stroma was isolated and cultured as described.³³

PCR

To study the presence of mRNA for both SDF-1 and its high affinity receptor CXCR4, cell pellets were made from confluent cultures of human BMEC (4LHBMEC, TrHBMEC and HBMEC60), HUVEC from two different donors, primary human bone marrow stroma and suspension cell lines KG-1a and KG-1v. From all samples RNA was isolated using Rneasy spincolumns (Qiagen, Hilden, Germany). cDNA synthesis was performed in a total volume of 40 l containing 1 RT buffer (Gibco), 0.1 mM DTT, 1 mM of each dNTP, 100 g/ml pdN6 (Pharmacia, Uppsala, Sweden), 1 U RNAsin (HT Biotechnology Ltd), 300 U M-MLV RT (Gibco) and 1 g RNA. RNA samples were heated for 5 min at 70°C before adding the cDNA synthesis mix and incubated for 2 h at 37°C. Five l of cDNA was used as a template in the PCR mix (1 PCR buffer (Gibco), 0.2 mM of each dNTP, 1.9 mM MgCl₂, 10 pmol of CXCR4 or SDF-1 forward and reverse primers, 0.5 U Platinum Taq polymerase (Gibco)). PCR was performed in a Hybaid PCR processor (15 s 99°C, 5 min 94°C (initially), 30 s 94°C, 30 s 56°C, 30 s 72°C during 35 cycles, followed by 10 min 72°C and 5 min 28°C). To verify proper RNA isolation and reverse transcription, a parallel PCR was performed on each sample using primers specific for the nonrearranged RAR transcripts of the retinoic acid receptor-alpha (RAR) gene.³⁴ RAR-specific fragments were amplified using 1 PCR buffer, 0.4 mM of each dNTP, 2 mM MgCl₂, 10 pmol of each RAR primer, 0.5 U Platinum Taq polymerase in one PCR round (15 s 99°C, 5 min 94°C (initially), 1 min 94°C, 1 min 50°C, 1 min 72°C during 35 cycles, followed by 10 min 72°C and 5 min 28°C. PCR products were electrophoresed on a 2% agarose gel. Primer sequences used were SDF-1 (forward) 5'-GTG GTC GTG CTG GTC CTC-3', (reverse) 5'-ATC TGA AGG GCA CAG TTT GG-3' (primer selection based on GenBank GI: 1220363); CXCR4 (forward) 5'-GTT ACC ATG GAG GGG ATC AG-3', (reverse) 5'-CAG CCT GTA CTT GTC CGT CA-3' (GI: 297099); RAR (forward) 5'-CCA GCT TCC AGT TAG TGG ATA TAG C-3', (reverse) 5'-ACC CCA TAG TGG TAG CCT GAG G-3' (GI: 4506418).

Flow cytometry and detection of endothelial SDF-1 binding via GAGs

In all flow cytometry experiments 100 000 cells were incubated with the primary antibody or appropriate isotype control for 30 min on ice. If necessary, a second step was performed with PE-labeled rabbit anti-mouse-Ig for another 30 min on ice. All cells were stained with 10% 7-amino-actinomycin D (Via-Probe; PharMingen, San Diego, CA, USA) to exclude dead cells. After each incubation cells were washed twice in PBS/0.1% BSA with 0.05% sodium azide. Cells were analyzed on a FACSCalibur flow cytometer (Becton Dickinson, San Jose, CA, USA) and data analysis was performed using the Cellquest software program (Becton Dickinson). To study involvement of GAGs in SDF-1 binding, 4LHBMEC or HUVEC were detached from culture plates with 10 mM EDTA, washed in PBS/0.1% BSA and 200 000–400 000 cells were incubated with medium alone or with specific GAG-degrading enzymes in medium 199 with 10 mM HEPES and 2 mM CaCl₂ for 60 min at 37°C. Enzymes used were: heparinases (a cocktail of 16 mU/ml heparinase (EC 4.2.2.7), heparitinase I (EC 4.2.2.8) and heparitinase II (EC 4.2.2.8) (Seikagaku, Tokyo, Japan), 1 U/ml chondroitinase ABC (Sigma) or trypsin (5 min incubation). Furthermore, cells were treated for 24 h with 60 mM sodium chlorate (Sigma) to inhibit incorporation of sulfate into GAGs. Cells were washed and incubated with 2.4 g/ml SDF-1 (R&D Systems, Abingdon, UK) in RPMI with 10 mM HEPES and 1% BSA or medium alone for 90 min at 4°C on a roller bank. Cells were transferred to tubes and were washed in PBS/0.1% BSA with 0.05% sodium azide. Cells were incubated with the anti-SDF-1 K15C antibody and second step labeling was performed as described above. After the second antibody step, cells were washed, fixed in 1% paraformaldehyde and analyzed. Mean fluorescence index (MFI) was calculated by the following formula: (mean fluorescence of the indicated antibody) - (mean fluorescence of the isotype control) / (mean fluorescence of the isotype control).

Competition of ¹²⁵I-SDF-1 binding to 4LHBMEC with GAGs

Competition experiments were performed by studying the binding of ¹²⁵I-SDF-1 (New England Nuclear (NEN), Du Pont, Wilmington, DE, USA; 2200 Ci/mmol) to 4LHBMEC in the presence of different GAGs as described earlier.²⁸ After detachment from culture plates with EDTA, 200 000 endothelial cells were incubated in 50 l RPMI/0.05% BSA with 0.5 ng ¹²⁵I-SDF-1 for 2 h on ice in the presence of various GAGs at a concentration of 200 g/ml or with medium alone. Cells were centrifuged twice at 700 g at 4°C for 5 min and cell pellets were resuspended in 500 l RPMI/0.05% BSA. Radioactivity was counted using a Wizard 140 automatic gamma counter. Experiments were performed in duplicate.

Isolation of PGs and HS-GAGs from 4LHBMEC and HUVEC

PGs were isolated from 4LHBMEC by a method previously described,²⁸ with some modifications. In brief, confluent cultures of 4LHBMEC and HUVEC in T175 culture flasks were metabolically labeled with 22 uCi/ml ³H glucosamine (NEN, 48.3 Ci/mmol) for 16 h at 37°C in MEM (Gibco) with 2% human serum. The medium was removed and after washing the cell layers were extracted with buffer containing PBS, 10 mM EDTA, 6 M urea, 1% Triton-X-100, 0.15 M NaCl, 8.5 mM Na₂HPO₄, 1.5 mM KH₂PO₄,1 mg/ml BSA and a cocktail of protease inhibitors (Roche, Almere, The Netherlands). PGs were isolated by chromatography over DEAE-Sephacell columns (3 1.5 cm, Pharmacia) and desalted on PD-10 columns (Sephadex G25, Pharmacia). To obtain GAGs, PGs were treated with 0.75 M NaOH in the presence of 20 mM NaBH₄ for 1 h at 73°C to release GAGs from core proteins (alkaline -elimination), neutralized with 6 M HCl and desalted. To isolate HS-GAGs, GAGs were treated with chondroitinase ABC (1 U/ml) overnight at 37°C. The obtained CS/DS disaccharides were removed by Superose 12 HR gel chromatography (Pharmacia) and the isolated HS-GAGs were desalted against destilled H₂O with PD-10 columns and lyophilized. The purity of isolated HS-GAGs was checked by treating the isolated fractions with nitrous acid at pH 1.5 followed by Superose 12 HR chromatography.

Binding of isolated endothelial HS-GAGs and PGs to SDF-1

Two assays were performed to study binding affinity of SDF-1 to isolated HS-GAGs or intact PGs from 4LHBMEC and HUVEC cell extracts. In the filter binding assay 10 000 c.p.m. of isolated ³H-labeled HS-GAGs from the cell extracts of 4LHBMEC and HUVEC, ³H-heparin or ³H-HS bovine kidney were incubated with 1 g SDF-1 (Baleux, Paris, France or R&D Systems). Chemokine-bound polysaccharides were trapped on a nitrocellulose filter. As a control basic fibroblast growth factor (bFGF) (Peprotech, Rocky Hill, NJ, USA) was used. Increasing concentrations of NaCl were added to release the ³H-labeled GAGs from the heparin/HS-SDF-1 complex. Radioactivity of the eluate was measured after addition of 10 ml Optiphase Hisafe 3 (Wallac, Breda, The Netherlands) in a liquid scintillation counter. In the microplate binding assay 10 g/ml SDF-1 (R&D Systems) or bFGF was immobilized on the bottom of microplates (96-wells, Costar) overnight at RT. Wells were blocked with PBS/BSA 1% for 2 h at RT and 10 000 c.p.m. of ³H-labeled total PGs (HS and CS/DS PGs) or ³H-heparin were added to the wells and incubated overnight. Wells were washed twice with PBS/BSA 0.1% to remove non-bound radioactivity and the bound GAGs were released by adding 2 M NaCl and radioactivity was measured as described above.

Flow adhesion experiments

Adhesion of HPC cell lines KG-1 and KG-1a to 4LHBMEC was measured under conditions of flow in a parallel plate flow perfusion chamber with well-defined rheological characteristics.^35,36 Endothelial cells were cultured to confluence in 3 or 4 days on gelatin-coated glass slides (18 18 mm) for use in the flow chamber. Twenty-four hours prior to perfusions the endothelial monolayer was washed with PBS and culture medium was replaced by medium without endothelial cell growth factor and heparin. Endothelial cells were stimulated with TNF (500 U/ml; Boehringer, Mannheim, Germany) 4–6 h prior to perfusions to upregulate adhesion molecules E-selectin, VCAM-1 and ICAM-1 (not shown). KG-1 or KG-1a cells were resuspended in HEPES flow buffer (20 mM HEPES, 132 mM NaCl, 6 mM KCl, 1.2 mM K₂HPO₄, 1 mM MgSO₄ 7H₂0, 5 mM glucose, 1 mM CaCl₂ at pH 7.4) at a concentration of 2 10⁶/ml and prewarmed to 37°C. The cells were aspirated from a reservoir through the perfusion chamber with a withdrawal syringe pump (sp210iw; World precision instruments, Sarasota, FL, USA). The flow chamber was mounted on a microscope stage (Zeiss) that was equipped with a video camera coupled to a VHS video recorder. After 1 min perfusion with HEPES flow buffer, cells were allowed to interact with the endothelium for 100 s at a shear stress of 1.0 dyn/cm², which is physiological in bone marrow microvessels in vivo.⁷ Thereafter, the supply of cells was stopped and exchanged for HEPES flow buffer and the shear stress was increased stepwise every 20 s to distinguish firmly arrested from loosely attached cells. The cells were subjected during 20 s to respectively 1.5, 2.0, 2.5, 3.5, 4.5, 5.5, 7.5 and 9.5 dyn/cm². Video recordings were made continuously and analyzed afterwards. One field of view (0.14 mm²) was evaluated during the increasing shear for the number of cells interacting with the endothelium at every time interval. The individual cells interacting with the endothelium before the incline in shear stress were followed during time; ie cells remained bound at the same position or were detached. After the last increment to 9.5 dyn/cm², the total number of firmly arrested cells was measured at a minimum of 30 randomized high power fields. All analyses were performed with the computer software program Optimas 6.1 (Media Cybernetics Systems, Silver Spring, MD, USA). Some slides with endothelial cells were pretreated for 5–20 min with 2 g/ml SDF-1 in HEPES flow buffer at RT and washed twice in flow buffer (37°C) before the perfusions. Preincubation of 4LHBMEC with a cocktail of heparinases (16 mU/ml) and chondroitinase ABC (1 U/ml) in medium 199 with 10 mM HEPES, 2 mM CaCl₂ and 0.5% BSA was performed for 60 min prior to perfusions to cleave both HS-GAGs and CS/DS-GAGs, respectively. Perfusions were performed in triplicate on at least four separate occasions.

Statistics

Results are expressed as mean standard error of the mean (s.e.m.). Statistical significance was determined with the Wilcoxon matched-pairs signed-ranks test or the Mann–Whitney U test with SPSS software (SPSS, Chicago, IL, USA).

Results

Expression of SDF-1 and CXCR4 by BMEC and HUVEC

We first established the expression levels of SDF-1 and its high affinity receptor CXCR4 on both mRNA and cell surface protein levels of microvascular BMEC and the macrovascular endothelial cell type HUVEC. In all samples, control PCR showed expression of the RAR gene, indicating the integrity of RNA and proper quality of the cDNA procedure (Figure 1a). Transcripts of SDF-1 (161 bp) were detected in the BMEC cell line 4LHBMEC (Figure 1a, lane 1) and two other bone marrow-derived endothelial cell lines: TrHBMEC and HBMEC60 (lanes 3, and 4). Stimulation of 4LHBMEC with TNF for 4 h did not seem to have an apparent effect on SDF-1 mRNA expression (lane 2). HUVEC from two different donors (lanes 5 and 6) and primary human bone marrow stromal cells (lane 7) all showed expression of SDF-1 mRNA. CXCR4 mRNA (240 bp) could be detected in all endothelial cell types and in bone marrow stromal cells (Figure 1a). As negative controls for SDF-1 expression both the myeloblastic cell line KG-1a (lane 9) and KG-1v (lane 8) were used, whereas for CXCR4 expression only KG-1a was negative.

Figure 1.

mRNA and cell surface expression of SDF-1 and CXCR4 by BMEC cell lines and HUVEC. (a) mRNA expression of SDF-1, CXCR4 and housekeeping gene RAR (RAR): 4LHBMEC without (lane 1) and with (lane 2) TNF stimulation (500 U/ml, four h). Lane 3: TrHBMEC; lane 4: HBMEC60; lanes 5 and 6: HUVEC from two donors; lane 7: human bone marrow stroma; lane 8: KG-1v; lane 9: KG-1a; lane 10: H₂O control. (b) Cell surface expression of SDF-1 (antibody MAB310) and CXCR4 on 4LHBMEC, HUVEC and KG-1v cells (positive control). Representative flow cytometry histograms are depicted indicating fluorescence staining. Dotted lines represent isotype controls.

Full figure and legend (41K)

With flow cytometry we studied the membrane expression of SDF-1 and CXCR4 on 4LHBMEC in comparison with HUVEC. EDTA-detached endothelial cells were incubated with two anti-SDF-1 antibodies: MAB310 and K15C. No SDF-1 could be detected with the MAB310 antibody on both endothelial cell types (Figure 1b), not even after addition of exogenous SDF-1 (not shown). In contrast, KG-1v cells expressed high levels of SDF-1 (Figure 1b). Because KG-1v displayed no SDF-1 mRNA it seems likely that the bound SDF-1 was derived from the serum in the culture medium. Recently, Amara et al²⁶ have described the anti-SDF-1 antibody K15C, which recognizes the first residues of SDF-1, which are involved in CXCR4 binding and exposed when SDF-1 is bound via HS-GAGs. Interestingly, using this antibody, SDF-1 could be detected on the cell surface of 4LHBMEC (Figure 2a). Binding of this antibody increased after incubation of 4LHBMEC with 2.4 g/ml SDF-1. SDF-1 binding was almost four-fold higher for 4LHBMEC compared to HUVEC (Figure 2b). To determine if 4LHBMEC and HUVEC expressed the SDF-1 receptor CXCR4 that could also be involved in SDF-1 binding, EDTA-detached cells were incubated with the PE-labeled CXCR4 antibody 12G5. However, in our hands no expression of CXCR4 on endothelial cells could be detected (Figure 1b), whereas it has been described by others that endothelial cells can express CXCR4.³⁷ TNF stimulation of 4LHBMEC or HUVEC (4 h) did not induce CXCR4 expression (not shown). As a positive control we used the KG-1v cell line which expresses high levels of CXCR4. EDTA treatment of these cells did not reduce CXCR4 expression (not shown), indicating that this was not the cause for lack of CXCR4 detection on the endothelial cells. Because recognition of bound SDF-1 with the K15C antibody and absence of protein levels of CXCR4 on 4LHBMEC were observed, these results strongly suggest a GAG-mediated binding of SDF-1. In summary, these experiments show that 4LHBMEC produce SDF-1 and express and bind SDF-1 on their cell surface. HUVEC also shows expression of SDF-1 mRNA, however, SDF-1 binds less efficiently as compared to 4LHBMEC.

Figure 2.

Detection of SDF-1 binding to 4LHBMEC and HUVEC with the K15C antibody. (a) 4LHBMEC was detached by EDTA and incubated with the anti-SDF-1 antibody K15C and rabbit anti-mouse-Ig-PE after incubation for 90 min with (+) or without (-) 2.4 g/ml SDF-1. The dotted line represents the isotype control. (b) 4LHBMEC and HUVEC were detached with EDTA, incubated with 2.4 g/ml SDF-1 and stained with the anti-SDF-1 antibody K15C and rabbit anti-mouse-Ig-PE. The MFI, representing the amount of bound SDF-1, of 4LHBMEC is fixed at 100% and relative binding is shown for the MFI of HUVEC. Mean percentage of binding is depicted from four paired experiments. *P = 0.014 (Wilcoxon matched-pairs signed-ranks test).

Full figure and legend (24K)

HS/heparin GAGs and CS/DS GAGs inhibit binding of SDF-1 to 4LHBMEC

To further study the involvement of GAGs in SDF-1 binding, various GAGs were investigated for their ability to compete with ¹²⁵I-SDF-1 binding to detached 4LHBMEC. Significant inhibition of SDF-1 binding was found for the low molecular weight heparin dalteparin (70.7 3.8% inhibition), unfractionated heparin (58.0 13.5%), HS intestinal mucosa (54.0 8.3%), CS (54.4 11.9%) and DS (43.0 18.5%) (Figure 3). The simultaneous addition of two or three competitors (heparin or dalteparin with CS, DS or both) did not further increase the inhibition of ¹²⁵I-SDF-1 binding (not shown). Surprisingly, addition of HS bovine kidney, an HS with a different sulfation pattern compared to HS intestinal mucosa, did not result in significant inhibition of SDF-1 binding. Preincubation of 4LHBMEC with a blocking anti-CXCR4 antibody (40 g/ml) did not affect binding of ¹²⁵I-SDF-1 (not shown). Thus, both heparin/HS, and CS/DS GAGs were able to inhibit SDF-1 binding to 4LHBMEC to a large extent.

Figure 3.

Competition of various GAGs with ¹²⁵I-SDF-1 binding to 4LHBMEC. Confluent cultures of 4LHBMEC were detached with EDTA, washed and 200 000 cells were incubated with 0.5 ng ¹²⁵I-SDF-1 in the presence of GAGs in a concentration of 200 g/ml for 2 h on ice. GAGs added were heparin, dalteparin, HS from bovine intestinal mucosa (HSIM), HS from bovine kidney (HSBK), chondroitin sulfate A (CS) and dermatan sulfate (DS). Radioactivity of cell pellets was counted after washing. Depicted is percentage binding compared to control situation (100% binding). Experiments were performed in duplicate on three separate occasions. *P < 0.04 compared to control; #P < 0.05 compared to competition with all other GAGs (Mann–Whitney U test).

Full figure and legend (14K)

SDF-1 binds to 4LHBMEC via GAGs

To confirm that SDF-1 binding to 4LHBMEC was mediated by GAGs, flow cytometry experiments were performed with the anti-SDF-1 K15C antibody after various pretreatment protocols. Trypsin treatment of 4LHBMEC completely abolished the binding of SDF-1 (Figure 4). Sodium chlorate treatment of the 4LHBMEC monolayer for 24 h, which inhibits incorporation of sulfate into GAGs,³⁸ reduced binding with 43.6 19.1%, indicating the requirement for the presence of sulfate groups in binding of SDF-1. Specific cleavage of the HS-GAGs on the cells with a cocktail of heparinases reduced binding of K15C with 53.9 7.7%. Degrading CS/DS GAGs with 1 U/ml chondroitinase ABC reduced 44.8 13.9% of binding, indicating not only SDF-1 mediated binding via HS-GAGs, but also via CS/DS-GAGs. Cleavage of both HS and CS/DS-GAGs resulted in 65.6 5.8% inhibition of SDF-1 binding, suggesting involvement of both receptors. These experiments show that HS and CS/DS-GAGs are the dominant binding sites of SDF-1 on the cell surface of 4LHBMEC.

Figure 4.

Binding of SDF-1 to cell surface GAGS on 4LHBMEC. 4LHBMEC were detached with EDTA and incubated with enzymes: trypsin, a cocktail of heparinases: heparinase, heparitinase I and heparitinase II (hepases), chondroitinase ABC (ABCase), both heparinases and chondroitinase ABC (hepases + ABCase) or pretreated with 60 mM sodium chlorate. After washing, the cell pellets were incubated with 2.4 g/ml SDF-1 for 90 min at 4°C. Subsequently, labeling with K15C and rabbit anti-mouse-Ig-PE was performed and fluorescence was measured with flow cytometry. The MFI, representing the amount of bound SDF-1 of treated cells relative to untreated cells (100% binding) is depicted. n = 3–4 separate experiments; *P = 0.014; #P = 0.037 compared to untreated cells (Mann–Whitney U test).

Full figure and legend (16K)

No binding of SDF-1 to isolated endothelial HS-GAGs and PGs

To study the binding affinity of SDF-1 to isolated endothelial GAGs, endothelial cell cultures were metabolically labeled with ³H glucosamine and total ³H-PGs and ³H-HS-GAGs were isolated from 4LHBMEC and HUVEC cell layers. Two different binding assays were performed: a filter binding assay³⁹ and a microplate binding assay with immobilized SDF-1. In the filter binding assay, isolated ³H-HS-GAGs, ³H-heparin or ³H-HS bovine kidney were incubated with 1 g/ml SDF-1 and subsequently trapped on nitrocellulose filters. In this assay, proteins, including protein–HS complexes, remain bound to the filter while non-bound radioactive ³H-HS-GAGs do not remain attached to the filter, but end up in the effluent. The concentration of NaCl used to release bound ³H-HS-GAGs from SDF-1 was plotted against release of radioactivity. SDF-1 bound weakly to ³H-heparin (Figure 5) and not at all to the different HS-GAGs. In contrast, and in accordance with literature,^40,41 the binding of the protein bFGF to heparin was more pronounced. To study whether intact PGs, ie GAGs attached to their core proteins, would have a better binding capacity towards SDF-1, SDF-1 was immobilized in a microplate and the total fraction of PGs (HSPGs and CS/DSPGs together (ratio approximately 1:1)²⁸ was allowed to bind. Similar results were obtained as in the filter binding assay, ie weak binding of heparin and no binding with isolated PGs, whereas in coated bFGF-wells both heparin and PGs were able to bind (not shown). These results show that soluble GAGs and PGs have a low binding affinity towards SDF-1.

Figure 5.

No SDF-1 binding to isolated HS-GAGs and PGs from 4LHBMEC and HUVEC. Filter binding assay. Isolated ³H-HS-GAGs from 4LHBMEC () and HUVEC cell-extracts () and ³H-HS bovine kidney (³H-HSBK) () were incubated with 1 g SDF-1 and ³H-heparin was incubated with either SDF-1 () or bFGF () and flushed over nitrocellulose filters. The decrease of the protein–GAG binding complex is depicted as cumulative percentual release of radioactivity (y-axis) and is plotted against the concentration of NaCl (x-axis) necessary to release the interaction.

Full figure and legend (22K)

Presented SDF-1 enhances arrest of KG-1 and not of KG-1a under flow conditions

To study the functional role of GAGs in presenting SDF-1, adhesion experiments were performed under flow conditions. To this end we used two HPC cell lines, KG-1, which expresses CXCR4⁺³² and KG-1a which does not express CXCR4 (Figure 1a). Functionality of the CXCR4 receptor in response to SDF-1 was first tested in a transmigration assay. KG-1 cells showed more migration towards an SDF-1 gradient compared to KG-1a cells (Figure 6a). In adhesion experiments under flow conditions, cells were perfused across slides cultured with confluent layers of TNF-stimulated 4LHBMEC at a physiological shear stress of 1 dyn/cm². Both cell lines express the E-selectin ligand cutaneous lymphocyte antigen (which enables these cells to roll over E-selectin) and integrins VLA-4 and LFA-1 (which are involved in firm adhesion to VCAM-1 and ICAM-1 respectively, not shown).

Figure 6.

Role of PG-mediated SDF-1 presentation in adhesion of KG-1 and KG-1a under flow conditions. (a) Transmigration of KG-1 and KG-1a towards SDF-1. 150 000 cells were allowed to migrate towards 180 ng/ml SDF-1 in a transwell system (24-well) for 4 h at 37°C. Inserts were coated with fibronectin. The percentage of transmigrated cells was calculated by flow cytometry in the presence of a known amount of flowcount beads (Beckman Coulter, Mijdrecht, The Netherlands) after correction for spontaneous migration (n = 3). (b) KG-1 (n = 13) or KG-1a (n = 4) were allowed to adhere to TNF stimulated 4LHBMEC for 100 s and then subsequently subjected to increasing shear stress every 20 s. Shown is the percentage of cells that were firmly adhered on the endothelium treated with SDF-1 (open bars) after the last increment in shear stress of 9.5 dyn/cm² compared to untreated endothelium (filled bars, 100% adhesion) measured from 30 high power fields. *P < 0.0001 compared to control (Mann–Whitney U test). (c) KG-1 cells were flowed over TNF stimulated 4LHBMEC treated or not with GAG-degrading enzymes (heparinases and chondroitinase ABC) with or without subsequent addition of 2 g/ml SDF-1 on the endothelium. The ratio of the number of cells interacting with the endothelium at 1.0 dyn/cm² which remained bound during the increasing shear stress is depicted compared to the control situation (ie no SDF-1 added and not treated with enzymes. The control represents 1.0 on the y-axis). n = 4–7 separate experiments; *P < 0.018 compared to control situation (Mann–Whitney U test).

Full figure and legend (27K)

After 1 min perfusion, 45.2 3.4% of KG-1 cells and 46.0 1.9% of KG-1a cells were rolling (n = 14, not shown). This rolling was largely E-selectin dependent for both cells (not shown). One hundred seconds after start of the perfusion experiments, the strength of adhesive cells was analyzed by measuring the resistance to detachment by controlled shear forces. Cells interacting with the endothelium were subjected to incrementing shear stress by increasing flow velocity every 20 s as explained in Materials and methods. After pretreatment of 4LHBMEC with SDF-1 (2 g/ml), more KG-1 cells remained bound during the increasing shear stress and consequently the number of firmly attached cells measured after the last increment to 9.5 dyn/cm² increased (from 127 16 to 208 30 cells/mm², Figure 6b). When KG-1a cells were used instead of KG-1 (without SDF-1 153 61 cells/mm² adhesion), no increase was observed after addition of SDF-1, indicating that SDF-1 acted via its receptor CXCR4 and not via an aspecific or indirect pro-adhesive effect on the endothelium (Figure 6b). When KG-1 cells interacting with the endothelium at 1.0 dyn/cm² were tracked during time, more cells remained bound at the same position compared to the control situation without SDF-1 added (3.4 times increase at the last increment to 9.5 dyn/cm², Figure 6c). This indicates that attached cells remained more resistant to high shear stress in the presence of SDF-1. To study the role of PGs in SDF-1 presentation, slides with 4LHBMEC were treated with a cocktail of GAG-degrading enzymes before adding SDF-1. Similar enzyme treatment abolished positive staining of 4LHBMEC monolayers with an anti-HS antibody (clone 10E4) using immunofluorescence microscopy (not shown). GAG degradation reduced the SDF-1-induced increase in firm arrest of KG-1 cells on SDF-1 treated 4LHBMEC especially at lower shear stress, although this was not significant (Figure 6c). Enzyme treatment alone did not differ significantly from the control situation. In conclusion, SDF-1 presented on 4LHBMEC increased the number of firmly adherent CXCR4⁺ KG-1 cells, which was partly due to proteoglycan-mediated SDF-1 binding.

Discussion

SDF-1 is a major homing factor for HPC.^17,23,42 To our knowledge this is the first report that demonstrates that PGs are the major SDF-1 binding site on BMEC. First, we showed by RT-PCR that SDF-1 is produced by the bone marrow-derived endothelial cell line 4LHBMEC and two other BMEC cell lines as well as by the macrovascular endothelial cell type HUVEC. This is in concordance with Imai et al,²⁵ who reported that murine BMEC cell lines expressed SDF-1 mRNA, whereas lung endothelial cells did not. Ponomaryov et al²⁴ however, could not detect SDF-1 mRNA in an endothelial cell line derived from bone marrow and HUVEC. This difference could be due to different cells used or to differences in sensitivity of the RT-PCR assays. We also showed that 4LHBMEC expressed low amounts of SDF-1 on the cell surface. Addition of SDF-1 resulted in extra binding to the cell membrane. Therefore, like others⁹ we added extra SDF-1 in all functional experiments to increase the observed effects.

Recently, we showed that radiolabeled SDF-1 could bind to 4LHBMEC and that this binding could be inhibited by adding O-sulfated heparin and to a lesser extent N-sulfated heparin as competitors.²⁸ These findings suggest the presence of an O-sulfated heparin-like receptor on 4LHBMEC to be involved in SDF-1 binding. In the present study we also show that other GAGs could inhibit the binding of SDF-1. Inhibition was found with heparin, low molecular weight heparin and HS from bovine intestinal mucosa origin as well as with CS and DS-GAGs. In contrast, no inhibition was seen with HS from bovine kidney origin. As both HS types do not differ much in overall sulfation, this suggests the absence of the SDF-1 binding motif in the polysaccharide chains from HS bovine kidney.

To further identify the SDF-1 binding site on 4LHBMEC, the anti-SDF-1 antibody K15C, which recognizes the CXCR4 binding site of SDF-1 was used.²⁶ This particular monoclonal antibody does not recognize SDF-1 when bound to CXCR4, but recognizes GAG-bound SDF-1. By using specific GAG-degrading enzymes we determined that HS-GAGs as well as CS/DS-GAGs mediated SDF-1 binding to endothelial cells. This is in contrast to Amara et al,²⁶ who could only detect HS-mediated SDF-1 binding on epithelial cells. Others, however, have also reported SDF-1 binding via CS chains.²⁷ Treatment with both GAG-degrading enzymes did not result in complete inhibition of SDF-1 binding (65.6%), suggesting the presence of (an)other unknown trypsin-sensitive receptor(s) for SDF-1 or incomplete cleavage by the enzymes used. The latter seems unlikely as for each enzyme activity was tested. Moreover, heparinases treatment induced high reactivity of 4LHBMEC with an antibody directed against enzyme-degraded HS-stubs (3G10)²⁸ (not shown). In addition to GAGs, mannose-containing species could be involved in SDF-1 binding.²⁷ Using another anti-SDF-1 antibody MAB310, no cell surface expression of SDF-1 could be detected on the endothelial cells. In contrast, the cell line KG-1v with high CXCR4 levels showed high SDF-1 expression with this antibody. These results might suggest that MAB310 recognizes an epitope in SDF-1 that is involved in polysaccharide binding and is exposed when SDF-1 is bound to CXCR4 and masked when SDF-1 is bound to polysaccharides.

In contrast to the binding of SDF-1 to cell membrane-bound PGs, no SDF-1 binding to isolated HS-GAGs or total PGs (both HS and CS/DS PGs) was seen in two different binding assays. SDF-1 bound weakly to heparin. These data indicate that soluble GAG-chains/PGs, in the low concentrations used in this assay, were not capable of binding SDF-1, in contrast to the control protein bFGF. It has been reported that SDF-1 applied on to a heparin-Sepharose column elutes at high ionic strength,²⁶ indicating a high affinity of SDF-1 for immobilized heparin. In general, carbohydrate–protein interactions are relatively weak. It could be that multivalency, the presence of many copies of oligosaccharides clustered in saccharide chains,^43,44 and/or a proper conformation of PGs and GAG chains on the cell surface are crucial for interaction with SDF-1. Recently, Zhang et al⁴⁵ reported that membrane-associated HSPGs but not similar soluble HSPGs induced FGFR signaling by bFGF. They suggested that membrane association of HSPGs is critical for interaction with the protein and that high density and cell surface association of PGs could impose particular chain orientation of GAGs resulting in optimal protein binding.

Sulfation patterns may be of critical importance in SDF-1 binding. We found that sulfation of the endothelium was involved in binding of SDF-1, as undersulfation of GAGs with sodium chlorate inhibited binding of SDF-1 with 43.6%. Recently, Gupta et al⁴⁶ reported that HS-GAGs derived from multi-potent non-hematopoietic stem cells from patients with Hurler disease (which results in abnormal GAG accumulation), differed in size and specific sulfation patterns from HS-GAGs of comparable normal stem cells. They described that these HS-GAGs indeed had differential binding capacities to SDF-1. Van der Voort et al⁴⁷ recently showed that cell surface HSPGs on activated B cells could bind hepatocytic growth factor but not SDF-1. This illustrates that HS is variable in structure and SDF-1 binding. In more detail, it has been proposed that 13 monosaccharides of heparin are involved in SDF-1 binding and that both 2-O and N-sulfation groups play a dominant role in this interaction.⁴⁸

HS-GAGs interact with many proteins. Especially, the binding of antithrombin to heparin has been the object of extensive research. The exact sulfation pattern of a heparin-derived pentasaccharide has been characterized, which accounts for the binding to antithrombin.⁴⁹ Furthermore, FGF and IL-8 are described to bind in a specific manner to HS-GAGs and heparin.^40,50 Kuschert et al⁵¹ reported that the chemokines IL-8, MCP-1, MIP-1 and RANTES bound to endothelial GAGs, with a higher requirement for O-sulfation than N-sulfation. As chemokines are able to bind preferentially to subsets of HS-GAGs,⁵² tissue-specific variation in the structure of GAGs could be responsible for the localization of particular chemokines to distinct regions of the vascular endothelium.⁵³ We showed previously that 4LHBMEC differed from HUVEC with respect to the amount of cell surface HS and HS sulfation patterns. We reported that radiolabeled SDF-1 bound better to 4LHBMEC compared to HUVEC,²⁸ although this increase was small. We now report almost four-fold less SDF-1 binding to HUVEC compared to 4LHBMEC using the anti-SDF-1 antibody K15C after preincubation of the cells with SDF-1. The difference between the results of antibody and ¹²⁵I-SDF-1 binding to endothelial cells might be that radiolabeled SDF-1 binds to all receptors, including possibly present high affinity receptors, whereas the anti-SDF-1 antibody recognizing the CXCR4-binding site is more specific for oligosaccharide-mediated binding. The difference between 4LHBMEC and HUVEC with respect to SDF-1 binding could be due to different GAGs and sulfation patterns of these GAG chains expressed by the endothelial cells.

 The immobilization of a chemokine as SDF-1 to endothelial GAGs can have multiple advantages compared to the presence of a soluble chemokine as also described by Tanaka et al.⁵⁴ First, the chemokine gets accumulated in a high concentration on the vascular wall to encounter target cells instead of being washed away by blood flow.⁵⁵ Second, through binding by PGs the CXCR4-binding site of SDF-1 is presented towards cells, as this site is distinct from the heparin-binding site.²⁶ Furthermore, by binding to a GAG chain, the configuration of the chemokine can change,⁵⁶ which optimizes the presentation of the chemokine to its receptor. Sadir et al.⁴⁸ showed that SDF-1 dimerizes when bound to heparin. This dimerization could enhance binding and signaling via CXCR4. Finally, by binding of the chemokine to GAGs, it will be protected from chemical and physiological degradation. It is likely that in vivo in the bone marrow endothelium newly synthesized HSPGs complex to SDF-1 in the Golgi apparatus. The PG holding the chemokine will then be transferred to the cell surface⁵⁷ and SDF-1 is then presented in the vascular lumen to encounting HPC.

Activation of integrins on HPC is essential for integrin-mediated adhesion in which a signal transduced to the HPC converts the functionally inactive integrin to an active adhesive configuration. SDF-1 has been shown to be able to induce signaling in HPC, resulting in adhesive integrins.^9,12 We showed that GAGs present on 4LHBMEC participate in SDF-1-mediated firm adhesion of HPC under flow conditions. When SDF-1 was preincubated on the endothelium, a statistical significant increase in firm arrest of KG-1 cells was observed, whereas this was not found for the CXCR4^- KG-1a cell line. These experiments support the hypothesis that by binding via PGs, SDF-1 is presented towards rolling HPC. The physiological relevance of SDF-1 binding by endothelial PGs could be the presentation of the CXCR4 binding site of SDF-1, which optimizes interaction with the receptor and thus enhances signaling events leading to integrin activation. Degradation of endothelial GAGs diminished the SDF-1-induced increase in firm adhesion of KG-1 cells, although a clear effect of SDF-1 could still be observed. This again indicates that some SDF-1 is able to bind to the endothelium after GAG degradation, possibly via other receptors as was also suggested in the flow cytometry binding experiments. Without the addition of exogenous SDF-1, firm adhesion was also observed, indicating also that a low amount of SDF-1 leads to arrest or that other mechanisms were also involved in this model. Integrins involved in SDF-1-induced firm adhesion could be VLA-4 integrins as KG-1 cells also firmly arrested on slides coated with VCAM-1 chimeras when SDF-1 was added to the substrate (not shown). As ICAM-1 is highly expressed on TNF-stimulated 4LHBMEC, 2-integrins are other candidates to be involved. Others have shown or proposed the role of endothelial PGs in chemokine presentation under flow conditions.^53,58 This is the first report, however, which describes the role of PGs on BMEC with regard to SDF-1 binding and presentation and subsequent enhanced arrest of HPC. In a recent paper we described the role of extracellular matrix HSPGs produced by BMEC in presenting SDF-1 and guiding migration of HPC.³² All together, bone marrow endothelial PGs could therefore play an important role in the homing of HPC to the bone marrow.

In summary, we have shown that cell surface HS and CS/DS PGs on bone marrow endothelium bind SDF-1 in a sulfate-dependent manner and present its CXCR4-binding site towards HPC thereby promoting adhesion of HPC to BMEC under flow conditions.

References

1. Rafii, S, Mohle, R, Shapiro, F, Frey, BM & Moore, MARegulation of hematopoiesis by microvascular endothelium. Leuk Lymphoma, (1997). 27, 375–386. | PubMed | ISI | ChemPort |
2. Vermeulen, M, Le Pesteur, F, Gagnerault, MC, Mary, JY, Sainteny, F & Lepault, FRole of adhesion molecules in the homing and mobilization of murine hematopoietic stem and progenitor cells. Blood, (1998). 92, 894–900. | PubMed | ISI | ChemPort |
3. Springer, TATraffic signals for lymphocyte recirculation and leukocyte emigration: the multistep paradigm. Cell, (1994). 76, 301–314. | Article | PubMed | ISI | ChemPort |
4. Butcher, EC & Picker, LJLymphocyte homing and homeostasis. Science, (1996). 272, 60–66. | Article | PubMed | ISI | ChemPort |
5. Mohle, R, Rafii, S & Moore, MAThe role of endothelium in the regulation of hematopoietic stem cell migration. Stem Cells, (1998). 16, 159–165. | PubMed |
6. Frenette, PS, Subbarao, S, Mazo, IB, von Andrian, UH & Wagner, DDEndothelial selectins and vascular cell adhesion molecule-1 promote hematopoietic progenitor homing to bone marrow. Proc Natl Acad Sci USA, (1998). 95, 14423–14428. | Article | PubMed | ChemPort |
7. Mazo, IB, Gutierrez-Ramos, JC, Frenette, PS, Hynes, RO, Wagner, DD & von Andrian, UHHematopoietic progenitor cell rolling in bone marrow microvessels: parallel contributions by endothelial selectins and vascular cell adhesion molecule 1. J Exp Med, (1998). 188, 465–474. | Article | PubMed | ISI | ChemPort |
8. Schweitzer, CM, Dräger, AM, Thijsen, SF, Zevenbergen, A, Theijsmeijer, AP, van der Schoot, CE & Langenhuijsen, MMConstitutive expression of E-selectin and vascular cell adhesion molecule-1 on endothelial cells of hematopoietic tissues. Am J Pathol, (1996). 148, 165–175. | PubMed | ISI | ChemPort |
9. Peled, A, Kollet, O, Ponomaryov, T, Petit, I, Franitza, S, Grabovsky, V, Slav, MM, Nagler, A, Lider, O, Alon, R, Zipori, D & Lapidot, TThe chemokine SDF-1 activates the integrins LFA-1, VLA-4, and VLA-5 on immature human CD34(+) cells: role in transendothelial/stromal migration and engraftment of NOD/SCID mice. Blood, (2000). 95, 3289–3296. | PubMed | ISI | ChemPort |
10. Schweitzer, CM, Vicart, P, Delouis, C, Paulin Dräger, AM, Langenhuijsen, MM & Weksler, BBCharacterization of a newly established human bone marrow endothelial cell line: distinct adhesive properties for hematopoietic progenitors compared with human umbilical vein endothelial cells. Lab Invest, (1997). 76, 25–36. | PubMed | ChemPort |
11. Rood, PM, Gerritsen, WR, Kramer, D, Ranzijn, C, von dem Borne, AE & van der Schoot, CEAdhesion of hematopoietic progenitor cells to human bone marrow or umbilical vein derived endothelial cell lines: a comparison. Exp Hematol, (1999). 27, 1306–1314. | Article | PubMed | ISI | ChemPort |
12. Peled, A, Grabovsky, V, Habler, L, Sandbank, J, Arenzana-Seisdedos, F, Petit, I, Ben-Hur, H & Lapidot, TThe chemokine SDF-1 stimulates integrin-mediated arrest of CD34(+) cells on vascular endothelium under shear flow. J Clin Invest, (1999). 104, 1199–1211. | PubMed | ISI | ChemPort |
13. Bleul, CC, Fuhlbrigge, RC, Casasnovas, JM, Aiuti, A & Springer, TAA highly efficacious lymphocyte chemoattractant, stromal cell-derived factor 1 (SDF-1). J Exp Med, (1996). 184, 1101–1109. | Article | PubMed | ISI | ChemPort |
14. Kawabata, K, Ujikawa, M, Egawa, T, Kawamoto, H, Tachibana, K, Iizasa, H, Katsura, Y, Kishimoto, T & Nagasawa, TA cell-autonomous requirement for CXCR4 in long-term lymphoid and myeloid reconstitution. Proc Natl Acad Sci USA, (1999). 96, 5663–5667. | Article | PubMed | ChemPort |
15. Pablos, JL, Amara, A, Bouloc, A, Santiago, B, Caruz, A, Galindo, M, Delaunay, T, Virelizier, JL & Arenzana-Seisdedos, FStromal-cell derived factor is expressed by dendritic cells and endothelium in human skin. Am J Pathol, (1999). 155, 1577–1586. | PubMed | ISI | ChemPort |
16. Aiuti, A, Webb, IJ, Bleul, C, Springer, T & Gutierrez-Ramos, JCThe chemokine SDF-1 is a chemoattractant for human CD34+ hematopoietic progenitor cells and provides a new mechanism to explain the mobilization of CD34+ progenitors to peripheral blood. J Exp Med, (1997). 185, 111–120. | Article | PubMed | ISI | ChemPort |
17. Peled, A, Petit, I, Kollet, O, Magid, M, Ponomaryov, T, Byk, T, Nagler, A, Ben-Hur, H, Many, A, Shultz, L, Lider, O, Alon, R, Zipori, D & Lapidot, TDependence of human stem cell engraftment and repopulation of NOD/SCID mice on CXCR4. Science, (1999). 283, 845–848. | Article | PubMed | ISI | ChemPort |
18. Mohle, R, Bautz, F, Rafii, S, Moore, MA, Brugger, W & Kanz, LThe chemokine receptor CXCR-4 is expressed on CD34+ hematopoietic progenitors and leukemic cells and mediates transendothelial migration induced by stromal cell-derived factor-1. Blood, (1998). 91, 4523–4530. | PubMed | ISI | ChemPort |
19. Mohle, R, Failenschmid, C, Bautz, F & Kanz, LOverexpression of the chemokine receptor CXCR4 in B cell chronic lymphocytic leukemia is associated with increased functional response to stromal cell-derived factor-1 (SDF-1). Leukemia, (1999). 13, 1954–1959. | Article | PubMed | ISI | ChemPort |
20. Oberlin, E, Amara, A, Bachelerie, F, Bessia, C, Virelizier, JL, Arenzana-Seisdedos, F, Schwartz, O, Heard, JM, Clark-Lewis, I, Legler, DF, Loetscher, M, Baggiolini, M & Moser, BThe CXC chemokine SDF-1 is the ligand for LESTR/fusin and prevents infection by T-cell-line-adapted HIV-1. Nature, (1996). 382, 833–835. | Article | PubMed | ISI | ChemPort |
21. Ma, Q, Jones, D, Borghesani, PR, Segal, RA, Nagasawa, T, Kishimoto, T, Bronson, RT & Springer, TAImpaired B-lymphopoiesis, myelopoiesis, and derailed cerebellar neuron migration in CXCR4- and SDF-1-deficient mice. Proc Natl Acad Sci USA, (1998). 95, 9448–9453. | Article | PubMed | ChemPort |
22. Zou, YR, Kottmann, AH, Kuroda, M, Taniuchi, I & Littman, DRFunction of the chemokine receptor CXCR4 in haematopoiesis and in cerebellar development. Nature, (1998). 393, 595–599. | Article | PubMed | ISI | ChemPort |
23. Kollet, O, Spiegel, A, Peled, A, Petit, I, Byk, T, Hershkoviz, R, Guetta, E, Barkai, G, Nagler, A & Lapidot, TRapid and efficient homing of human CD34(+)CD38(-/low)CXCR4(+) stem and progenitor cells to the bone marrow and spleen of NOD/SCID and NOD/SCID/B2m(null) mice. Blood, (2001). 97, 3283–3291. | Article | PubMed | ISI | ChemPort |
24. Ponomaryov, T, Peled, A, Petit, I, Taichman, RS, Habler, L, Sandbank, J, Arenzana-Seisdedos, F, Magerus, A, Caruz, A, Fujii, N, Nagler, A, Lahav, M, Szyper-Kravitz, M, Zipori, D & Lapidot, TInduction of the chemokine stromal-derived factor-1 following DNA damage improves human stem cell function. J Clin Invest, (2000). 106, 1331–1339. | PubMed | ISI | ChemPort |
25. Imai, K, Kobayashi, M, Wang, J, Shinobu, N, Yoshida, H, Hamada, J, Shindo, M, Higashino, F, Tanaka, J, Asaka, M & Hosokawa, MSelective secretion of chemoattractants for haemopoietic progenitor cells by bone marrow endothelial cells: a possible role in homing of haemopoietic progenitor cells to bone marrow. Br J Haematol, (1999). 106, 905–911. | Article | PubMed | ChemPort |
26. Amara, A, Lorthioir, O, Valenzuela, A, Magerus, A, Thelen, M, Montes, M, Virelizier, JL, Delepierre, M, Baleux, F, Lortat-Jacob, H & Arenzana-Seisdedos, FStromal cell-derived factor-1alpha associates with heparan sulfates through the first beta-strand of the chemokine. J Biol Chem, (1999). 274, 23916–23925. | Article | PubMed | ISI | ChemPort |
27. Mbemba, E, Gluckman, JC & Gattegno, LGlycan and glycosaminoglycan binding properties of stromal cell-derived factor (SDF)-1alpha. Glycobiology, (2000). 10, 21–29.
28. Netelenbos, T, Dräger, AM, Van het Hof, B, Kessler, FL, Delouis, C, Huijgens, PC, van den Born, J & Van Dijk, WDifferences in sulfation patterns of heparan sulfate derived from human bone marrow and umbilical vein endothelial cells. Exp Hematol, (2001). 29, 884–893. | PubMed | ChemPort |
29. Höök, M, Riesenfeld, J & Lindahl, UN-[3H]Acetyl-labeling, a convenient method for radiolabeling of glycosaminoglycans. Anal Biochem, (1982). 119, 236–245. | Article | PubMed | ISI |
30. Rood, PM, Calafat, J, von dem, B, Gerritsen, WR & van der Schoot, CEImmortalisation of human bone marrow endothelial cells: characterisation of new cell lines. Eur J Clin Invest, (2000). 30, 618–629. | Article | PubMed | ISI | ChemPort |
31. Jaffe, EA, Nachman, RL, Becker, CG & Minick, CRCulture of human endothelial cells derived from umbilical veins. Identification by morphologic and immunologic criteria. J Clin Invest, (1973). 52, 2745–2756. | PubMed | ISI | ChemPort |
32. Netelenbos, T, Zuijderduijn, S, van den Born, J, Kessler, F, Zweegman, S, Huijgens, PC & Dräger, AMProteoglycans guide SDF-1-induced migration of hematopoietic progenitor cells. J Leukoc Biol, (2002). 72, 353–362. | PubMed |
33. Zweegman, S, Veenhof, MA, Debili, N, Schuurhuis, GJ, Huijgens, PC & Dräger, AMMegakaryocytic differentiation of human progenitor cells is negatively influenced by direct contact with stroma. Leukemia, (1999). 13, 935–943. | Article |
34. Jansen, JH, de Ridder, MC, Geertsma, WM, Erpelinck, CA, van Lom, K, Smit, EM, Slater, R, vd Reijden, BA, de Greef, GE, Sonneveld, P & Lowenberg, BComplete remission of t(11;17) positive acute promyelocytic leukemia induced by all-trans retinoic acid and granulocyte colony-stimulating factor. Blood, (1999). 94, 39–45. | PubMed | ISI | ChemPort |
35. Sakariassen, KS, Aarts, PA, de Groot, PG, Houdijk, WP & Sixma, JJA perfusion chamber developed to investigate platelet interaction in flowing blood with human vessel wall cells, their extracellular matrix, and purified components. J Lab Clin Med, (1983). 102, 522–535. | PubMed | ISI | ChemPort |
36. Kuijper, PH, Gallardo, TH, van der Linden, JA, Lammers, JW, Sixma, JJ, Koenderman, & Zwaginga, JJPlatelet-dependent primary hemostasis promotes selectin- and integrin-mediated neutrophil adhesion to damaged endothelium under flow conditions. Blood, (1996). 87, 3271–3281. | PubMed | ISI | ChemPort |
37. Gupta, SK, Lysko, PG, Pillarisetti, K, Ohlstein, E & Stadel, JMChemokine receptors in human endothelial cells. Functional expression of CXCR4 and its transcriptional regulation by inflammatory cytokines. J Biol Chem, (1998). 273, 4282–4287. | Article | PubMed | ISI | ChemPort |
38. Humphries, DE & Silbert, JEChlorate: a reversible inhibitor of proteoglycan sulfation. Biochem Biophys Res Commun, (1988). 154, 365–371. | Article | PubMed | ChemPort |
39. Maccarana, M & Lindahl, UMode of interaction between platelet factor 4 and heparin. Glycobiology, (1993). 3, 271–277. | Article | PubMed | ISI | ChemPort |
40. Maccarana, M, Casu, B & Lindahl, UMinimal sequence in heparin/heparan sulfate required for binding of basic fibroblast growth factor. J Biol Chem, (1993). 268, 23898–23905. | PubMed | ISI | ChemPort |
41. Turnbull, JE, Fernig, DG, Ke, Y, Wilkinson, MC & Gallagher, JTIdentification of the basic fibroblast growth factor binding sequence in fibroblast heparan sulfate. J Biol Chem, (1992). 267, 10337–10341. | PubMed | ISI | ChemPort |
42. Onai, N, Zhang, Y, Yoneyama, H, Kitamura, T, Ishikawa, S & Matsushima, KImpairment of lymphopoiesis and myelopoiesis in mice reconstituted with bone marrow-hematopoietic progenitor cells expressing SDF-1-intrakine. Blood, (2000). 96, 2074–2080. | PubMed | ISI | ChemPort |
43. Carver, JP, Michnick, SW, Imberty, A & Cumming, DAOligosaccharide-protein interactions: a three-dimensional view. Ciba Found Symp, (1989). 145, 6–18.
44. Roseman, DS & Baenziger, JUThe mannose/N-acetylgalactosamine-4-SO4 receptor displays greater specificity for multivalent than monovalent ligands. J Biol Chem, (2001). 276, 17052–17057.
45. Zhang, Z, Coomans, C & David, GMembrane heparan sulfate proteoglycan-supported FGF2-FGFR1 signaling. Evidence in support of the cooperative-end-structures model. J Biol Chem, (2001). 276, 41921–41929. | Article | PubMed | ISI | ChemPort |
46. Gupta, P, Brazil, JJ, Reyes, M & Verfaillie, CMHuman bone marrow derived mesodermal progenitor cells (MPC) in vitro correct the biochemical abnormality in Hurler syndrome. Blood, (2000). 96, 279a
47. Van der Voort, R, Taher, TE, Wielenga, VJ, Spaargaren, M, Prevo, R, Smit, L, David, G, Hartmann, G, Gherardi, E & Pals, STHeparan sulfate-modified CD44 promotes hepatocyte growth factor/scatter factor-induced signal transduction through the receptor tyrosine kinase c-Met. J Biol Chem, (1999). 274, 6499–6506. | Article | PubMed | ISI | ChemPort |
48. Sadir, R, Baleux, F, Grosdidier, A, Imberty, A & Lortat-Jacob, HCharacterization of the stromal cell-derived factor-1alpha-heparin complex. J Biol Chem, (2001). 276, 8288–8296. | Article | PubMed | ISI | ChemPort |
49. Lindahl, UWhat else can 'Heparin' do?. Haemostasis, (1999). 29, 38–47.
50. Spillmann, D, Witt, D & Lindahl, UDefining the interleukin-8-binding domain of heparan sulfate. J Biol Chem, (1998). 273, 15487–15493. | Article | PubMed | ISI | ChemPort |
51. Kuschert, GS, Coulin, F, Power, CA, Proudfoot, AE, Hubbard, RE, Hoogewerf, AJ & Wells, TNGlycosaminoglycans interact selectively with chemokines and modulate receptor binding and cellular responses. Biochemistry, (1999). 38, 12959–12968. | Article | PubMed | ISI | ChemPort |
52. Witt, DP & Lander, ADDifferential binding of chemokines to glycosaminoglycan subpopulations. Curr Biol, (1994). 4, 394–400. | Article | PubMed | ISI | ChemPort |
53. Tanaka, Y, Adams, DH & Shaw, SProteoglycans on endothelial cells present adhesion-inducing cytokines to leukocytes. Immunol Today, (1993). 14, 111–115. | Article | PubMed | ISI | ChemPort |
54. Tanaka, Y, Kimata, K, Adams, DH & Eto, SModulation of cytokine function by heparan sulfate proteoglycans: sophisticated models for the regulation of cellular responses to cytokines. Proc Assoc Am Physicians, (1998). 110, 118–125. | PubMed | ISI | ChemPort |
55. Butcher, ECLeukocyte–endothelial cell recognition: three (or more) steps to specificity and diversity. Cell, (1991). 67, 1033–1036. | Article | PubMed | ISI | ChemPort |
56. Hoogewerf, AJ, Kuschert, GS, Proudfoot, AE, Borlat, F, Clark-Lewis, I, Power, CA & Wells, TNGlycosaminoglycans mediate cell surface oligomerization of chemokines. Biochemistry, (1997). 36, 13570–13578. | Article | PubMed | ISI | ChemPort |
57. Templeton, DMProteoglycans in cell regulation. Crit Rev Clin Lab Sci, (1992). 29, 141–184. | PubMed |
58. Weber, KS, von Hundelshausen, P, Clark-Lewis, I, Weber, PC & Weber, CDifferential immobilization and hierarchical involvement of chemokines in monocyte arrest and transmigration on inflamed endothelium in shear flow. Eur J Immunol, (1999). 29, 700–712. | Article | PubMed | ISI | ChemPort |

Acknowledgements

The anti-SDF-1 antibody K15C was a generous gift from F Arenzana-Seisdedos (Unité d'Immunologie Virale, Paris, France). The authors thank Jan van der Linden (Department of Pulmonary Diseases, Utrecht Medical Center, Utrecht, The Netherlands) for providing us with the Optimas 6.1 software and Suzanne Zuijderduijn for performing transmigration experiments.