Synthetic heparan sulfate dodecasaccharides reveal single sulfation site interconverts CXCL8 and CXCL12 chemokine biology† †Electronic supplementary information (ESI) available. See DOI: 10.1039/c5cc05222j

Multigram-scale synthesis of a sulfation-site programmed dodecasaccharide is described. CXCL8- and CXCL12-mediated in vitro and in vivo biology is shown to be regulated by a single sulfation site change.


Materials and Methods 1 Synthetic experimental and data General methods
All the chemicals used were purchased from commercial sources without further purification. All reactions were monitored by TLC on Merck silica gel plates 60 F254. Silica gel 60 (particle size 0.035-0.070 mm) was used for column chromatography. 1 H NMR spectra were recorded at 800 or 400 MHz and 13 C spectra at 100 or 200 MHz respectively on Bruker DPX spectrometers. Mass spectra (MS) were recorded using a Micromass Platform II spectrometer using an electro spray ionization source or via the EPSRC National Mass Spectrometry Service (Swansea). Isotope patterns for compounds with mass > 1000 are included in the SI. Optical rotations were obtained using an AA-1000 polarimeter. Elemental analyses were performed by Micro Analytical Laboratory, School of Chemistry, The University of Manchester. NMR Data were reprocessed using iNMR 4 from Nucleomatica and Bruker Topspin.

HS competition assay.
HS competition assays were performed as previously described 2 . Briefly, ninety six-well heparinbinding plates (Iduron) were coated with HS (Iduron). After blocking and washes, 10 ng of IL-8 or SDF-1α (both from R&D Systems) were diluted in 100 µL of standard assay buffer (50 mM sodium acetate pH 7.3, 150 mM sodium chloride and 0.2% Tween20) containing 10 mg/mL BSA. Oligosaccharides were mixed at 0.1-100 µg/mL range for both IL-8 and SDF-1α and added to immobilized HS for 2 hours. After washes, HRP-tagged primary antibodies against IL-8 (R&D Systems, Quantikine ELISA kit) or SDF-1α (R&D Systems, Quantikine ELISA kit) were added to the wells for 2 hours. Following the washes, TMB solution (Sigma-Aldrich) was added for 30 min and the reaction was stopped with 2M sulphuric acid. Optical density was measured at 450 nm.

In vitro assays for testing angiogenic functions of endothelial cells.
HUVEC migration in the presence or absence of oligosaccharides was tested using a wound healing assay 2 . Briefly, 2X10 4 HUVECs were seeded per well of a 96-well plate and incubated at +37 º C, in 5% CO2. Confluent monolayers were serum-starved in EBM-2 medium lacking supplements containing 0.1% FBS for 24 hours. Cell monolayers were wounded with a pipette tip and overlayed with EBM-2 medium without supplements with 0.1% FBS or medium supplemented with IL-8 (50 ng/mL) or SDF1-α (50 ng/mL; both from R&D Systems). Synthetic oligosaccharides were administered at 1, 10 and 50 µg/ml. Phase contrast images were taken immediately after beginning the treatment and after 24 hours using Zeiss Axiovert 200M microscope (Zeiss, Hertfordshire, UK) enclosed in a full environmental chamber. The unpopulated areas at the beginning and the end of each experiment were measured using MetaMorph image analysis software (Molecular Devices). The cell advancement area was derived for each treatment.

Chemotaxis assay.
The transwell migration assay was performed using 24-well transwell units with 8 µm pore size PVDF polycarbonate uncoated filters (Costar). MCF10A-CXCR4 cells (5x10 4 ) were seeded in serum-free DMEM/F12 in the upper compartment of the chemotaxis chambers. DMEM/F12 supplemented with 5% horse serum containing either recombinant SDF-1α (100 ng/mL; R&D Systems) or PBS as control was added to the lower well to stimulate directional migration of MCF10A-CXCR4 cells. Oligosaccharides were used at 50 µg/mL concentration. MCF10A-CXCR4 cells were allowed to migrate for 24 h at 37 ºC in a tissue culture incubator. The cells remaining on the upper surface of the membrane were removed, and those present on the lower surface of the filter were fixed in 4% PFA for 5 min, stained with DAPI and counted at x10 magnification using Olympus BX51 microscope. Twenty independent fields per filter were analysed.

Gelfoam in vivo angiogenesis assay.
All procedures were carried out in accordance with UKCCCR guidelines 1999 by approved protocol (Home Office Project license no. 40-3609). Sterile absorbable sponges (Pharmacia, Peapack, NJ) were cut into 5 × 5 × 7-mm pieces and hydrated overnight at 4 °C in sterile PBS. Excess PBS was then adsorbed onto sterile filter paper. The sponges were then soaked with 0.4% agarose (100 µl) containing either saline, 250 ng of CXCL8 (R&D Systems), 250 ng of CXCL12 (PreproTech) or chemokines plus each oligosaccharide (250 µg). The agarose-Gelfoam plugs were then allowed to harden for 1 hour at room temperature before s.c. implantation into mice. Female SCID-bg mice (Paterson Institute, The University of Manchester) were anesthetized with isoflurane, and the plugs were implanted s.c. via a midline incision of the abdominal skin and placed either toward the right flank or the left flank. Each group consisted of 6 mice. After seven days gelfoam plugs were harvested, placed in OCT solution, and snap-frozen in liquid nitrogen for subsequent immunohistochemical analyses. The sections were stained with anti-mouse CD31 as described 3 . Monoclonal anti-α-SMA (clone 1A4) Cy3 conjugate (Sigma-Aldrich) diluted 1:50 was added to visualize mural cells. Neutrophils were visualized with anti-Ly-6B.2 alloantigen antibody (clone 7/4; AbD Serotec) diluted 1:100 and incubated overnight at 4 ºC. AlexaFluor488-conjugated donkey anti-rat IgG antibody (1:1000; Invitrogen) was used as a secondary antibody. Macrophage specific anti-receptor F4/80 rat monoclonal antibody (AbD Serotec) was used at 1:1000 dilution and secondary AlexaFluor488-conjugated donkey anti-rat antibody (Invitrogen) was used at 1:1000. Images were collected with 3D-Histech Mirax scanner (3DHISTECH), viewed with Pannoramic Viewer software (3DHISTECH) and analysed by ImageJ software for CD31-positive blood vessels, neutrophil and macrophage staining. The whole area in each sponge section was evaluated using Metamorph image analysis software (Molecular Devices). The number of blood vessels in each analysed section was normalized to the standardized section area.

Leukocyte Transmigration Assay
Leukocyte transmigration through HUVEC monolayer in response to chemokines CXCL12 CXCL8 and CCL19 (MIP-3β) was tested using the CytoSelect Leukocyte Transmigration Assay (Cell Biolabs, Cambridge, UK) following the manufacturer's protocol.
Briefly, 7x10 5 HUVEC in EBM-2 medium containing all supplements were added to each insert, which was placed in a 24-well plate containing 500 µl of complete EBM-2 medium. HUVEC were cultured for 30 hours following addition of 40 ng/ml of TNFα for a further 18 hours.
Twenty ml of peripheral blood was collected from healthy volunteers and diluted with 20 ml of HBSS lacking calcium and magnesium (Fisher Scientific, Loughborough, UK). Fifteen ml of Ficoll-Paque Plus solution (GE Healthcare, Uppsala, Sweden) was added. After centrifugation at 400 g for 35 minutes at room temperature, buffy coats were collected, aliquoted and frozen for long-term storage in liquid nitrogen.
When required, buffy coats were cultured for 24 hours in RPMI medium containing glutamine and supplemented with 10% FBS. Cells were collected by centrifugation at 400 g for 5 minutes, resuspended at 1x10 6 /ml and LeukoTracker staining solution was added at 1:500 for 1 hour. Cells were washed 3 times with RPMI containing 0.5% FBS and resuspended in RPMI containing 10% FBS. The inserts were removed from EBM-2 medium and placed in RPMI 10% FBS that contained CXCL12, CXCL8 or CCL19 at 1 µg/ml concentration or a mix of each chemokine and respective oligosaccharide at 1 mg/ml concentration. Oligosaccharide and chemokines were pre-incubated for 2 hours before adding to the inserts containing a HUVEC monolayer. Leukocytes (3x10 5 ) were added to the upper chamber and maintained there for 2 hours in tissue culture incubator. Transmigrated leukocytes were lysed in solution and fluorescence was measured at 480/520nm.

Statistical analysis.
Data are expressed as the mean ±SEM. For comparison of groups, the two-tailed Student's t test was used. A level of P < 0.05 was considered as statistically significant.          (a) 12--mer ISNSm6S showed the greatest potency in inhibiting CXCL12--induced ERK phosphorylation. Serum--starved HUVECs were stimulated with 50 ng/mL of CXCL12 for 10 minutes in the absence or presence of oligosaccharides. Stimulation by CXCL12 alone was taken as 100%. Average values derived from four independent experiments are expressed as the mean ±SEM. Average values for treatment with a heparin 12--mer were derived from two independent experiments and expressed as the mean ±SD. (b) oligosaccharide ISNS is a more potent inhibitor of CXCL8 than CXCL12. IL--8 (50 ng/mL) stimulation of serum--starved HUVECs was performed for 10 minutes in the absence or presence of respective oligosaccharides. Phosphorylation of STAT3 was detected with the antibody recognizing phosphorylated serine 727. Stimulation with CXCL8 alone was deZined as 100%. Average values derived from densitometric evaluation of four independent experiments were expressed as the mean ±SEM. The average effect of a heparin 12--mer was derived from two independent experiments and expressed as the mean ±SD.   The effects of dodecasaccharides on CXCL--8 and CXCL12--induced HUVEC migration were tested in a wound healing assay. The impact of cytokines on wound healing without oligosaccharides is expressed as 100%. Oligosaccharides 3 and dp12 were added at 50 µg/ml. The values are derived from two independent experiments performed in triplicate and represent the mean ± SD. *, p 0.006; ‡, p 0.05. dp12 contains a terminal C=C in place of the polar 4--OH group. The synthetic 4--OS systems reported here retain the terminal ring conformation but the close comparability of dp12 and synthetic per--O--6 sulfated 3 conZirms that the 4--OS is not a signiZicant variable in any of the effects seen here.

31
CXCL8 CXCL12 dp12 dp12 Supplementary Figure 26: Statistical data for Fig.2c and additional data covering oligosaccharide 3. Table 1. P--values generated by comparing the means of CXCL12 and CXCL8 binding to HS without oligosaccharides (controls) and in the presence of 100 µg/ml of each oligosaccharide. Table 2. P--values generated for the inhibiHon of IL--8 and SDF--1α binding to HS by ISNS and ISNSmono6S oligosaccharides. Table 3. P--values generated by comparing the mean values of IL--8 and SDF--1α binding to HS in the presence of ISNS6S oligosaccharide with the values of ISNS or mono6S oligosaccharide treatment. Table 4. P--values generated for the inhibiHon of IL--8 and SDF--1α binding to HS by heparin dp12 and ISNS or ISNSmono6S oligosaccharides.