Glycan materials: amino-terminating glycan probes
AEAB-terminating92 high-mannose N-glycan probes #7-#12 were purchased from NatGlycan (Atlanta, Georgia, USA). Aminopropyl-terminating probes #6, #91-110, #170, #174-#177, and #187, synthesised chemically, were purchased from GlycoNZ (Auckland, New Zealand). The following amino-terminating glycans were obtained as gifts from different resources or collaborators: Chemically synthesised aminopropyl (Sp)-terminating probes #155 and #156 were from the Core D of the US Consortium for Functional Glycomics; probes #165 and #169 were from Prof. Nicolai Bovin (Ovchinnikov Institute of Bioorganic Chemistry, Russian Academy of Sciences), probe #18293 from Prof Chunxia Li (Ocean University of China, Qingdao, China), and GalNAc β-linked to Ser (#163) and Thr (#164)93 were from Ulrika Westerlind (Umeå University). The glycine (Gly)-terminating glycan probes #79-#81, #84, #151-#154, #157-159, #160-162, and #183-186 are N-glycosides and were prepared by conjugation of the respective reducing sugars with glycine as described94, and purified further by HPLC. All amino-terminating glycans were analyzed by electrospray (ESI) or matrix-assisted laser desorption/ionisation (MALDI) mass spectrometry (MS) (Source Data File 1) with hexose-containing glycans quantified by micro-scale hexose assay (see below) before microarray printing.
Preparation and analysis of amino-terminating glycan probes
Deprotection of Fmoc-protected and per-acetylated serine- and threonine-linked mucin cores
Probes #1-#3 and #166-#168 were prepared by deprotection of fluorenylmethoxycarbonyl (Fmoc)- and per-acetyl Ser- and Thr-linked mucin cores purchased from Sussex Research Laboratories (Ottawa, Canada). In brief, 1 N NaOCH3 in methanol ( ~ pH 10-11) was added to the solution of the protected sample (1 mg in 0.5 ml methanol). The solution was stirred for 3 h at room temperature and the pH was adjusted to pH 4-5 using acetic acid. The sample solution was dried under a N2 stream and re-dissolved in 0.05 M ammonium acetate before purification by gel filtration on a Superdex Peptide column (Pharmacia, Uppsala, Sweden) with elution by 0.05 M ammonium acetate. The collected fraction was freeze-dried, and ammonium acetate removed by repeated co-evaporation with water. The deprotected samples were analyzed by MS and quantified by micro-scale hexose assay (see below) before use.
Extended and Ser-linked mucin core 3 O-glycans
Enzymatically prepared Ser-terminating probes #178-#180 were from Prof Kelly Moremen (Complex Carbohydrate Research Center, University of Georgia) and were purified by gel filtration chromatography using Superdex Peptide column before use.
Ser-terminating probe #181 was chemically synthesised and as a gift from Prof Dr Akihiro Imamura of Gifu University (Yanagido, Japan). These Ser-linked and extended mucin core 3 O-glycans were analyzed by ESI-MS (Fig. 1 in Source Data File 1) and quantified by hexose assay before microarray printing.
Preparation of amino-terminating probes from azido-terminating glycans by click chemistry
Probes #19-#22, #27, #28, #30-39, #41, #43, #44, #48, #50, #51, #54, #56, #57, #61-#63, #65-#67, #69, #70, #72-#77 were prepared from chemoenzymatically synthesised 3-azidopropyl glycans95 after conversion into amino-terminating glycans. Conversion of the azido- into amino-terminating glycans was performed as described previously96 with the following modifications. To the freeze-dried mixture of azido-glycans (50 nmol) and triethylene glycol 2-aminoethyl propargyl ether (50 nmol, Sigma-Aldrich, Gillingham, England) was added pre-mixed 200 ml solution of t-butyl alcohol/water (2:1, v/v) containing CuSO4 (30 nmol), sodium ascorbate (30 nmol) and tris-benzyltriazolylmethyl amine (TBTA, 60 nmol). The mixture was incubated at room temperature for 1 h with occasional vortexing. The solvent was removed under reduced pressure and the residue purified by gel filtration on a Superdex Peptide column as described above. Further purification was carried out by HPLC on a XBridge amide column (Waters, Wilmslow, England) with a binary solvent system of A: acetonitrile/H2O 70:30 and B: acetonitrile /H2O 20:80, with both containing 10 mM KH2PO4 (pH 3.0). The elution was carried out by gradient 5-25%B in 45 min at a flow rate of 1 ml/min and detection by UV 206 nm. The collected fractions were desalted by gel filtration on a Superdex Peptide column as described above. The converted propyl triazole-PEG4-amine (PTPA)-terminating probes were purified by HPLC (Fig. 2 in Source Data File 1) for representative HPLC profiles) and analyzed by MALDI-MS (Table 1 in Source Data File 1 for representative mass spectra). Quantitation was by micro-scale hexose assay (see below) before microarray printing.
Preparation of AEAB-terminating probes from reducing glycans by reductive-amination
Neutral glycan (#29, #40, #42, #45-#47, #49, #52, #53, #55, #58-#60, #64, #68, #71 and #82) and sialylated probes (#16, #17, #23-#25), purchased respectively from Elicityl (Crolles, France) or Dextra Laboratories (Reading, England), were prepared from reducing sugars whereas sulfated KS oligosaccharide probes #126-#135, were prepared from chemically synthesised reducing sugars gifted by Tokyo Chemical Industry (Tokyo, Japan). The sialylated probes #18 and #26, were prepared from the enzymatic synthesised reducing sugars (Dr Apoorva Srivastava, Department of Chemistry, Imperial College London). Preparation of the fluorescent reagent AEAB and conjugation with reducing sugars by reductive-amination were performed as previously described92. For purification of the conjugation product, a Sep-Pak Aminopropyl cartridge (Waters, Wilmslow, England) was used with elution by acetonitrile /H2O for the neutral sugars and acetonitrile /0.05 M ammonium acetate for the sialylated sugars. Further purification was carried out by HPLC using Amide column (XBridge, Waters). The gradient elution was acetonitrile /H2O for the neutral and acetonitrile /0.05 M ammonium acetate for the sialylated sugars at a flow rate of 1 ml/min. with detection by UV 330 nm. The detection and probe quantitation were performed using by UV 330 nm as described92.
Glycosaminoglycan oligosaccharide-AEAB probes
Oligosaccharide fractions from heparin97, chondroitin sulphate A, B and C98, and keratan sulphate98 were prepared by partial digestion with heparin lyase I (Sigma), chondroitinase ABC (Sigma), and keratanase II, respectively, as previously described. Hyaluronic acid (HA) fragments were prepared in two different ways. HA lyase (EC 4.2.2.1; from Streptomyces hyalurolyticus; Sigma) was used to obtain unsaturated ΔUA-terminating oligosaccharides and HA hydrolase (EC 3.2.1.35; from bovine testes; Sigma) was used to prepare the unmodified HA oligosaccharides99. After conversion into AEAB probes, the conjugation products of KS DP10 and 14, (#136-#138), HA DP4, 8, 11-14 (#139-#146), and CS DP6, 10 and 14 (#111-#122) and non-sulphated chondroitin DP10, 12 and 14 (#88-#90), and the conjugation products of heparin DP6, 10 and 14 (#123-#125) were purified using an amino-cartridge (Sep-Pak Aminopropyl, Waters) with elution by sodium acetate (0.05 M). Additional purification was carried out by a short column of strong anion exchange (HiTrap Q, GE Healthcare, Amersham, England) with elution by a gradient of NaCl. The collected fractions were desalted on a Superdex Peptide column with elution by ammonium acetate (0.05 M).
Serine- and threonine-linked sialyl mucin core 1 probes
Mono- and disialyl core 1 probes #4, #5, #171-#173 with either Ser or Thr were isolated from human urine as described100. Further purification was carried out by HPLC with XBridge amide column (Waters, Wilmslow, England) with a binary solvent system of A: acetonitrile/H2O 70:30 and B: acetonitrile/H2O 20:80 (both containing 7.5 mM KH2PO4). The elution was carried out by a gradient 15-20%B in 40 min at a flow rate of 1 ml/min and detection by UV 206 nm.
Human milk oligosaccharide AEAB probes
AEAB probes #78 and #83-#87 were prepared from DP6-8 and DP10 fractions of HMOs as described101. Amide and PGC columns were used for separation and purification. Their sequences were determined by negative-ion ESI-CID-MS/MS98.
Complex-type N-glycan AEAB probes from fetuin N-glycans
Fully sialylated bi- and tri-antennary N-glycan AEAB probes #13-#15, were prepared from fetuin N-glycans released by PNGase F (New England Biolabs, Ipswich, MA) and isolated from a short column of Bio-Gel P6 as described102. The total released N-glycans were conjugated to AEAB and the access AEAB reagent was removed by NH2-cartridge. The N-glycan-AEAB fraction was fractionated by HPLC using a Hypersil APS-2 column (Fisher Scientific, Loughborough, England) with a gradient by acetonitrile/H2O containing ammonium acetate. Detection was by UV at 330 nm and the flow rate was at 1 ml/min.
Micro-scale hexose quantitation
The micro-scale dot orcinol assay on TLC plates was performed as previously described103. In brief, hexose-containing probes, especially those not AEAB conjugated, were dissolved in H2O (at approx. 0.5-1.0 mg/ml), and 1 μl was spotted onto a silica gel TLC plate. Standard solutions (Galactose in H2O) were also applied alongside the glycan probes. The plate was dried and sprayed with orcinol/H2SO4 staining reagent and heated in an oven at 105 oC to develop the colour. Quantitation was carried out by scanning at 550 nm on a CAMAG TLC scanner (Omicron Research, Hungerford, England).
MS analysis of amino-terminating glycan probes
MS analysis of most amino-terminating probes was carried out by an electrospray mass spectrometer on a Waters (Manchester, UK) Q-TOF-type mass spectrometer SYNAPT-G2 in either positive- or negative ion mode. Cone voltage was at 50 eV or 80 eV and capillary voltage at 3 kV. Source temperature was 80 oC and the desolvation temperature 150 oC. A scan rate of 1.5 sec/scan was used and the acquired spectra were summed for presentation. For analysis, glycan probes were dissolved in H2O typically at a concentration of 10–20 pmol/μl, of which 1 μl was injected via an HPLC injector. Solvent acetonitrile/2 mM ammonium bicarbonate 1:1 was delivered by a HPLC pump (Waters) at a flow rate of 10 μl/min. Selected probes were also analyzed by MALDI-MS on an Axima MALDI Resonance mass spectrometer with a QIT-TOF configuration (Shimadzu). A nitrogen laser was used to irradiate samples at 337 nm, with an average of 100-200 shots accumulated. A matrix solution (0.5 μl) of 2,5-dihydroxybenzoic acid (20 mg/ml) in a mixture of methanol/water (1:1) was deposited on the sample target before application of the sample solution (0.5 μl).
Biotinylated GAG polysaccharide probes
Biotinylated CSA, CSB, CSC and heparin (#147-#150, respectively) were prepared and purified essentially as described104. The level of biotin introduced was one for every 50 carboxyl groups105. In brief, 50 mg of GAG polysaccharide were dissolved in 0.1 M MES buffer (pH 5.5) before addition of 109 ml of 25 mM biotin LC-hydrazide (Pierce, Tattenhall, UK) solution in Me2SO and 130 ml of 25 mg/ml 1-ethyl-2-(2-dimethylamino propyl) carbodiimide solution in 0.1 M MES buffer. The reaction mixtures were stirred at room temperature for 24 h and dialysed extensively against deionized water. A short Sephadex G-10 column (1.6 × 30 cm) was used to remove remaining biotin and other reagents. Quantitation of GAG polysaccharides and their modified forms was carried out by carbazole assay106.
Bacterial culture and labelling
Bacterial isolates used in this study were commercially sourced (ATCC and DSMZ) or were kindly gifted by the Lawley and Wolfe lab (Supplementary Data 4). Isolates from the Wolfe lab were originally obtained from non-pregnant women who provided verbal and written consent following Loyola University Medical Centre Institutional Review Board approval as detailed32. The identity of the bacterial species used was confirmed using sanger sequencing targeting the 16S rRNA gene and MALDI_TOF.
Bacterial Culture
E. coli isolates were streaked on LB agar (Sigma, Lennox recipe) and incubated overnight at 37 °C. LB broth was inoculated from cultures on plates and incubated overnight in a shaking incubator at 37 °C and 200 rpm. Anaerobic bacteria were cultured in the following media: L. crispatus, MRS agar and broth media; L. iners and G. vaginalis, Columbia (Difco) + 5% v/v defibrinated horse blood (Oxoid) agar and Vaginal Microbe Media (VMM, Patent protected); S. agalactiae, BHI (Oxoid) agar and broth media; and F. nucleatum, Columbia (Difco) + 5% v/v Defibrinated horse blood (Oxoid) agar and BHI (Oxoid) broth supplemented with 5 g/l yeast extract (Sigma Aldrich), 0.005 g/l L-cysteine hydrochloride monohydrate (Merck), 0.01 g/l D-(+)-cellobiose (Sigma Aldrich) and 0.01 g/l D-(+)-maltose (0.01 g/l) media. Agar (1.2% w/v) was added to the media to prepare agar plates. Isolates of anaerobes were streaked on agar plates and incubated overnight at 37 °C in an anaerobic chamber (10% CO2, 10% hydrogen, 80% nitrogen and 70% humidity). A total of 10 ml of degassed broth was inoculated and incubated overnight in anaerobic conditions at 37 °C. The starter broth was used to inoculate the required volume of fresh media and was incubated as described above until the broth reached an OD600 ≈ 1.0.
Fluorescent labelling of cultured live bacteria
For staining of live bacteria, 2 ml aliquots of culture were centrifuged for 10 minutes, and the resulting pellet washed twice with 1 ml of cold sterile Hanks Balanced Saline solution (HBSS) (Gibco). 1 × 108 bacterial cells were stained with 10 μM CellTrace™ CFSE (Molecular Probes) or 10 μM CellTrace™ Far Red (Molecular Probes) in 250 uL of PBS and incubated for 1 h in a rotating wheel at room temperature protected from light. Bacteria were centrifuged, the pellet was washed twice with cold sterile HBSS and resuspended in 1 ml of cold sterile HBSS. Stained bacteria were kept on ice or at 4 °C and protected from light until further use (within 2 h for live bacteria).
Fluorescent labelling of fixed bacterial cultures
E. coli and F. nucleatum cultures were centrifuged for 10 minutes to pellet bacteria and cells were washed twice in cold sterile PBS. Bacteria were fixed with 4% formaldehyde in PBS in a rotating wheel at room temperature for 15 min. E. coli or F. nucleatum fixed cells were centrifuged and washed twice in cold sterile PBS, the cell pellet was resuspended in HBSS (250 uL for 1 × 108 bacterial cells) before staining with 10 μM SYTO™59 (Invitrogen) or 10 μM SYTO™ BC Green (Invitrogen) respectively, for 30 min in a rotating wheel at room temperature protected from light. Bacteria were centrifuged, the pellet was washed twice with sterile cold PBS and resuspended in the desired volume of cold sterile PBS. Stained bacteria were kept on ice or at 4 °C and protected from light until further use (ideally no more than 1 week).
All centrifugations were done at 5000 x g for 10 min in a refrigerated centrifuge set at 4 °C, except for E. coli where 3000 x g were used to avoid fimbriae shedding. HBSS media and buffers required for staining and fixing procedures of anaerobe bacterial cultures were degassed before use. Wide-bore tips were used while handling bacterial cultures to avoid fimbriae shedding. To ensure the bacterial cultures were not contaminated during culturing, we used chromogenic agar (CHROMagar™) and light microscopy to visualise homogeneous and distinct cell morphology.
Glycan microarray preparation and analyses
The generation of covalent sequence-defined glycan arrays and the GAG polysaccharide microarray on NHS-functionalised glass slides (Nexterion H) was performed essentially as described before49,98. Five oligosaccharide microarray sets and one GAG polysaccharide array set were generated (Supplementary Fig. 1 and Supplementary Data 1). Details of the preparation of the microarrays and the methods used for binding assays and data analysis are in accordance with MIRAGE (Minimum Information Required for A Glycomics Experiment) guidelines for reporting of glycan microarray-based data107 (Supplementary Data 3).
Binding analyses with anti-carbohydrate antibodies, plant lectins and other glycan binding proteins
Details of the antibodies, lectins and glycan binding proteins used are summarised in Supplementary Data 2. Microarray analyses were performed at ambient temperature essentially as previously described108. No blocking was used for overlay assays on covalent glycan array glass slides. Pre-complexation of hSigLec15-Fc with biotinylated anti-human IgG (1:1 by weight) was performed at 4 °C for 30 min before adding it to the microarray.
Bacterial binding analyses
Microarray analyses of fluorescently labelled bacteria (see above) was performed as follows. Fluorescently labelled bacterial cultures at OD = 1 were resuspended in binding buffer (HBS: 10 mM HEPES at pH 7.0 or pH 4, 150 mM NaCl, supplemented with 5 mM CaCl2 or Acetate buffer: 10 mM sodium acetate, at pH 7.0 or pH 4, 150 mM NaCl, supplemented with 5 mM CaCl2) and were used for overlays on glycan microarrays. 100 µL of bacterial suspension was applied to the incubation chamber with the microarrays, incubated for 1 h at room temperature under mild agitation on an oscillating platform and washed three times with binding buffer, followed by two washes with HPLC grade water to remove salts from the array. Slides were dried under a mild nitrogen flow and scanned for quantitation as described below. HBS at pH4 was prepared immediately before use by adjusting the pH of HBS with HCl. Acetate buffer pH4 was prepared immediately before use by adjusting the pH with Acetic acid. The pH of the bacterial suspensions was monitored before and after overlays using pH strips.
Microarray data analysis
Images of fluorescently bound antibodies, lectins, glycan binding proteins and bacteria were acquired with a GenePix 4300 A scanner from Molecular Devices, quantified using GenePix software and processed with CarbArryART109. Bar charts and heatmaps were generated using excel. RANK quantitation is described in MIRAGE (Supplementary Data 3) and performed as previously described.
Glycan binding on-array inhibition assays
For on-array inhibition studies of F. nucleatum 23726 glycan binding, fluorescently labelled cultures of bacteria were incubated with 10 mM D-galactose, 10 mM D-glucose, or HEPES buffer (HBS: 10 mM HEPES at pH 7.5, 150 mM NaCl, and 5 mM CaCl2), for 15 min at room temperature before performing bacterial overlay on glycan microarrays.
Bacterial competition assays on array
For on-array competition studies between L. crispatus and S. agalactiae, 50 μl of 0.2 × 108, 0.4 × 108, 0.8 × 108, or 1.6 × 108 cells/ml of CellTrace™ CFSE labelled L. crispatus cultures were added to 50 μl of 0.4 × 108 cells/ml CellTrace™ Far Red fluorescently labelled cultures of S. agalactiae (see staining procedures described above and MIRAGE (Supplementary Data 3) before performing bacterial overlay on glycan microarrays.
Mammalian cell culture
VK2/E6E7 vaginal epithelial cells (ATCC- CRL-2616) were cultured and maintained in Keratinocyte SFM 1X (Gibco) containing Calcium Chloride 0.4 mM and 1% w/v penicillin– streptomycin (Sigma Aldrich). Cells were cultured at 37 °C and 5% CO2 in a humidified incubator and were free of mycoplasma, which was routinely tested for using the MycoStrip Detection Kit (InvivoGen).
Chondroitinase and heparin lyase III digestion of VK2 cells
The VK2/E6E7 cells were gently scrapped and resuspended in PBS at a concentration of 8.105 cells/ml. Chondroitinase ABC (E.C. 4.2.2.4, Sigma) and heparin lyase III (E.C. 4.2.2.8, recombinant, IBEX Technologies, Montreal, Canada) were added at a concentration of 0,005 U/ml and 0,001 U/ml respectively and the cells were incubated at 37 °C for 2 h. The cells were centrifuged 5 min at 300 x g and resuspended in PBS at the appropriate pH for antibody or bacteria binding analyses.
Assessment of VK2-glycan and bacteria binding by flow cytometry
Sub-confluent cultures of VK2/E6E7 cells were gently scrapped and resuspended in PBS. Single cell suspensions were stained with anti-chondroitin sulphate (CS56, Abcam), anti-keratan sulphate (MZ15, a present from Fiona Watt, Imperial, London), anti-heparan sulphate (clone 10E4, Amsbio) or biotinylated hyaluronic acid binding protein (Millipore), followed by Alexa Fluor 488-conjugated anti-mIgG (ThermoFisher) or Alexa Fluor 647-conjugated anti-mIgM antibodies (ThermoFisher) or streptavidin PE (Phycoerythrin) conjugate (Miltenyibiotech), in PBS for 30 min at 4 °C. Cells were analyzed using a FACSCalibur flow cytometer (Becton Dickinson). Intact VK2 cells were gated based on their FFS and SSC, excluding cells with extreme low or high FSC and SSC values, see example in Supplementary Fig. 8). Flow cytometry data was analyzed using CellQuest software (BD Biosciences).
For bacterial binding assays, VK2/E6E7 cells were added to bacteria suspensions to obtain a ratio of bacterial particles to cells or MOI (Multiplicity of Infection)=20 for all bacterial strains except for G. vaginalis at pH 4 where MOI = 100 was used. Cell mixtures were gently stirred for 1 h at 4 °C. Where appropriate, fluorescently labelled bacteria were incubated in PBS containing polysaccharides at 5 mg/ml for 30 min before binding assays. Cells were analyzed using a FACSCalibur flow cytometer (Becton Dickinson) and analyzed using the CellQuest software (BD Biosciences).
Electron microscopy
Overnight cultures of E. coli isolates were imaged by negative stain. For staining, 10 µl of culture was deposited onto glow-discharged 300 mesh carbon-coated copper grids (Agar Scientific), and left for 2 min. The grid was washed twice with 10 µl of water. The sample was stained with 0.2% (w/v) or 0.04% (w/v) phosphotungstic acid for 7 sec, after which the stain was blotted off. The grid was imaged on a Tecnai 12 Spirit transmission electron microscope (FEI) operating at 120 kV and with a 2 K eagle camera (FEI).
SaHyal_767 protein expression and purification
The gene WP_000403395.1 encoding a putative hyaluronate lyase enzyme from S. agalactiae 767 was analyzed using InterProScan. The designed construct Sa_Hyal_767 includes both the carbohydrate-binding and catalytic domain (residues 106-1081). Tyr578 was mutated to Phe to reduce enzymatic activity on glycan binding assays and was codon optimised for E.coli expression. The construct was synthesised and cloned between the BamHI/XhoI sites in the multiple cloning site of the expression vector pET28a(+) (GenScript) (Supplementary Fig. 10). The resulting recombinant protein has a hexa-histidine-tag at the N-terminal. Protein expression was carried out in E. coli strain BL21(DE3) induced by adding 1 mM isopropyl-β-d-thiogalactopyranoside, overnight at 19 °C. IMAC purification was performed with a imidazole gradient. Fractions containing the purified recombinant proteins were buffer exchanged to 50 mM HEPES, pH 7.5, 100 mM NaCl, 5 mM CaCl2 and 0.02% sodium azide. The final yield of purified protein was approximately 8 mg/L of culture.
Isothermal calorimetry (ITC) assays
The theoretical basis of kinetic measurements by ITC is described elsewhere110. Briefly, in a single injection ITC experiment, the total heat measured is proportional to the apparent enthalpy (ΔHapp), and the number of moles of product generated. The reaction rate can be related to the amount of heat generated over time. The affinity for the substrate (Km) and the turnover rate (kcat) values of the enzyme can be obtained by fitting the rate of the reaction versus the substrate concentration plots with a Michaelis-Menten curve. Experiments were performed on MicroCal PEAQ-ITC calorimeter (Malvern). One single 38 μl injection was performed with a speed of 0.5 μL s–1. The instrument was set to high-feedback mode, reference power 5 μcal sec-1, stirring speed of 650 rpm, and an experimental temperature of 25 °C. A 600 second pre-injection delay was applied for baseline stabilisation after equilibration. Apparent enthalpy of the reaction, heat rate (dQ/dt), and Michaelis–Menten plots were generated using MicroCal PEAQ-ITC Analysis Software (Malvern).
Sa_Hyal_767 activity against HA 14-mer, heparin 14-mer, and LNT was tested at pH 7.5. The enzyme was dialysed overnight into a buffer either containing 50 mM HEPES at pH 7.5, 150 mM NaCl, and 5 mM CaCl2. The lyophilised glycans were dissolved in the same corresponding buffer. A Sa_Hyal_767 solution at 50 nM was placed in the sample cell, and a 0.5 mM glycan solution was loaded in the injection syringe. Substrate autohydrolysis and heat of dilution phenomena were assessed by replacing the enzyme solution in the calorimetric cell with reaction buffer and carrying out the measurements with identical experimental settings.
Statistical analysis
In glycan microarray binding analyses each assay was performed with four technical replicates and results in graphs presented as mean +/− SD of quadruplicates (Source Data Files 2–7). Bacterial binding to polysaccharide GAG arrays was performed in three independent assays. Results in graphs are presented as mean +/− SD of quadruplicates on the microarray of representative experiments or as mean +/− SD of three independent experiments. Statistical significance of data from independent experiments was assessed using an unpaired two-tailed Student’s t test. Microscopy images are representative and similar results were observed in different fields of view with two independent biological replicates. Mammalian cell-based assays were performed with at least three biological replicates. Results in graphs are presented as mean +/− SD. Statistical significance was assessed using 2-way ANOVA.
Reporting summary
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.
