Bronchial mucins from patients suffering from CF (cystic fibrosis) exhibit glycosylation alterations, especially increased amounts of the sialyl-Lewisx (NeuAcα2-3Galβ1-4[Fucα1-3]GlcNAc-R) and 6-sulfo-sialyl-Lewisx (NeuAcα2-3Galβ1-4[Fucα1-3][SO3H-6]GlcNAc-R) terminal structures. These epitopes are preferential receptors for Pseudomonas aeruginosa, the bacteria responsible for the chronicity of airway infection and involved in the morbidity and early death of CF patients. However, these glycosylation changes cannot be directly linked to defects in CFTR (CF transmembrane conductance regulator) gene expression since cells that secrete airway mucins express no or very low amounts of the protein. Several studies have shown that inflammation may affect glycosylation and sulfation of various glycoproteins, including mucins. In the present study, we show that incubation of macroscopically healthy fragments of human bronchial mucosa with IL-6 (interleukin-6) or IL-8 results in a significant increase in the expression of α1,3/4-fucosyltransferases [FUT11 (fucosyltransferase 11 gene) and FUT3], α2-6- and α2,3-sialyltransferases [ST3GAL6 (α2,3-sialyltransferase 6 gene) and ST6GAL2 (α2,6-sialyltransferase 2 gene)] and GlcNAc-6-O-sulfotransferases [CHST4 (carbohydrate sulfotransferase 4 gene) and CHST6] mRNA. In parallel, the amounts of sialyl-Lewisx and 6-sulfo-sialyl-Lewisx epitopes at the periphery of high-molecular-mass proteins, including MUC4, were also increased. In conclusion, our results indicate that IL-6 and -8 may contribute to the increased levels of sialyl-Lewisx and 6-sulfo-sialyl-Lewisx epitopes on human airway mucins from patients with CF.
Human bronchial mucins consist of a family of polydisperse glycoproteins of high molecular mass, with different polypeptide chains (named apomucins) substituted with carbohydrate chains in serine- and threonine-rich tandem repeats that are responsible for approx. 70–80% of their apparent molecular mass . Apomucins of bronchial mucins are encoded by at least eight MUC (mucin) genes (MUC1, MUC2, MUC4, MUC5B, MUC5AC, MUC7, MUC8 and the recently characterized MUC19) . Carbohydrate chains that cover the apomucins are extremely diverse, adding to the complexity of these molecules. Currently, more than 150 different O-linked carbohydrate chains have been described, and it is assumed that the bronchial mucins from a single individual probably contain a few hundred different carbohydrate chains . In addition, because bronchial mucins are located at the surface of airways, where they have a protective function, the broad diversity of their carbohydrate peripheral determinants could be involved in multiple interactions with bacteria and viruses, which are then normally eliminated by the mucociliary system.
The biosynthesis of O-linked carbohydrate chains is a stepwise process involving a number of glycosyl- and sulfotransferases. The only structural element shared by all mucin O-glycan chains is a GalNAc residue α-linked to a serine or threonine residue of the apomucin. The non-reducing end of carbohydrate chains, which corresponds to the termination of the biosynthetic process, may bear different structures, such as blood groups A, B or O determinants; H and sulfated H determinants; Lewisa, Lewisb, Lewisy or various derivatives of the Lewisx epitope, such as the sialyl-Lewisx, 3-sulfo-Lewisx and 6-sulfo-sialyl-Lewisx determinants. The synthesis of these different terminal determinants involves different pathways utilizing a variety of transferases such as (i) α2,3-sialyltransferases [encoded by ST3GAL genes (α2,3-sialyltransferase genes)], (ii) α1,3-fucosyltransferases [FUT (fucosyltransferase gene)], (iii) galactose-3-O-sulfotransferases [GAL3ST (Gal-3-O-sulfotransferase gene)] and (iv) GlcNAc-6-O-sulfotransferases (CHST) . In the lung, it is not very clear yet which fucosyl-, sialyl- and sulfotransferases are involved in the biosynthesis of these determinants that are present at the periphery of mucins and probably crucial in many pathologies, including CF (cystic fibrosis).
Inflammation plays a major role in the pathophysiology of lung disease in CF: the inflammatory response is thought to be abnormal and/or excessive in CF, which would contribute to the degradation of the respiratory function of the patients. It is still a matter of debate whether there may or may not be a pre-inflammation of the lungs as part of the basic functional defect of the CFTR (CF transmembrane conductance regulator). Armstrong et al.  have suggested that inflammation may even precede infection in this disease. The presence of increased neutrophils and cytokines such as TNFα (tumour necrosis factor α), IL-6 (interleukin-6) and IL-8 in young infants with CF with no evidence of infection has also been shown [6,7]. There is also accumulating evidence that the secretion of several cytokines, especially IL-8, is up-regulated in CF . In contrast, bronchoalveolar lavages from CF patients contain less anti-inflammatory cytokine IL-10 than healthy controls and bronchial epithelial CF cells secrete less IL-10 than normal cells, suggesting that there could be an imbalance between pro- and anti-inflammatory cytokines in CF, leading to an exaggerated inflammatory response that would be at the onset of the lung disease and/or would contribute to the lung disease development [9,10].
Glycosylation defects on glycoconjugates secreted by CF patients or present on the membrane of CF cells have been described since the 1970s. Bronchial mucins secreted by CF patients are more sulfated, sialylated and fucosylated, and salivary and intestinal mucins from CF patients also contain more sulfate than healthy individuals [11–13]. In addition, it has been shown the sialyl-Lewisx epitope is overexpressed on bronchial mucins from severely infected CF patients, but also on bronchial mucins from infected patients suffering from other lung pathologies . 1H-NMR and mass spectroscopy have been used to determine the structures of oligosaccharides in bronchial mucins from CF and non-CF patients, which confirmed the increased presence of sialyl-Lewisx epitopes and HO3S-6-GlcNAc residues .
Interestingly, it seems that glycosylation of secreted and membrane glycoconjugates is either not affected by the level of expression of CFTR [16,17] or that membrane-bound CF glycoconjugates and secreted glycoconjugates are differently affected. CF membrane glycoconjugates show a decreased sialylation compared with non-CF , whereas mucins secreted by CF patients are oversialylated. These discrepancies suggest that factors other than the CFTR deficiency could be responsible for the altered glycosylation and sulfation of CF mucins. It is of particular interest to understand the origin of the sialylation, fucosylation and sulfation modifications carried by CF bronchial mucins, because the sialyl-Lewisx and 6-sulfo-sialyl-Lewisx epitopes that are overexpressed on these mucins are preferential receptors for Pseudomonas aeruginosa, the bacteria responsible for the progressive destruction of CF patients' lungs . As already mentioned, it has been shown that inflammation could induce modifications of glycosylation in the lung, such as an overexpression of the sialyl-Lewisx epitope on bronchial mucins . In addition, the pro-inflammatory cytokine TNFα can induce increased expression and activities of some sialyl-, fucosyl- and sulfotransferases in human bronchial explants as well as in the human respiratory glandular cell line (MM-39) . IL-6 and IL-8 are present in high amount in bronchoalveolar lavage fluids from CF, but their effect on glycosylation is still unknown.
In the present study, we have investigated the effect of these two cytokines on the expression and activity of bronchial glycosyl- and sulfotransferases. We show an increased expression of several fucosyl-, sialyl- and sulfotransferases, as well as increased amounts of sialyl-Lewisx and 6-sulfo-sialyl-Lewisx epitopes on bronchial proteins, including MUC4, which could influence the development of the lung pathology.
The mAbs (monoclonal antibodies) against the sialyl-Lewisx (CSLEX1) and 6-sulfo-sialyl-Lewisx (G152) epitopes were from BD Pharmingen (San Diego, CA, U.S.A.) and a gift from Dr R. Kannagi (Department of Molecular Pathology, Research Institute, Aichi Cancer Center, Japan) respectively. HRP (horseradish peroxidase)-labelled goat anti-mouse IgG secondary antibodies were obtained from Amersham Biosciences (Uppsala, Sweden). Cytokines IL-6 and IL-8 were from AbCys S.A. (Paris, France). Culture media were from Invitrogen (Paisley, Renfrewshire, Scotland, U.K.). Alkaline phosphatase-conjugated MAA (Maackia amurensis agglutinin) and SNA (Sambucus nigra agglutinin) were from E.Y. Laboratories (San Mateo, CA, U.S.A.). Clostridium perfringens sialidase was from Sigma.
In accordance with the Declaration of Helsinki (2000) of the World Medical Association, after ethics committee approval (CHRU Lille) and informed consent from each patient had been obtained, tissues were collected from macroscopically healthy areas of the bronchial tree of patients undergoing surgery for bronchial carcinoma. They were immersed in Leibovitz L15 medium (Invitrogen), immediately transported on ice to the laboratory and then processed to isolate the mucosa. Mucosa (1–2 cm2) was cut into 1 mm2 pieces and suspended in CMRL-1066 medium (Invitrogen) complemented with 0.2 mM L-glutamine . They were maintained at 37 °C for 16 h in the presence or absence (controls) of 20 ng/ml IL-6 or IL-8 (AbCys).
RNA isolation and cDNA synthesis
Total RNA was isolated from fragments of bronchial mucosa incubated with and without IL-6 or IL-8 for 16 h, using the NucleoSpin® RNA II (Macherey-Nagel). Isolated RNA (1–2 μg) was then subjected to reverse transcription in the presence of oligodeoxythymidilic acid12-18 primer (First strand cDNA synthesis kit; Amersham Biosciences) for cDNA synthesis in a final volume of 33 μl according to the manufacturer's instructions.
Semi-quantitative and quantitative PCR analysis of mucin, glycosyltransferase and sulfotransferase transcriptional expression
The oligonucleotides used as primers for the PCR reactions are given in Supplementary Tables 1 and 2 (available online as Supplementary data at http://www.BiochemJ.org/bj/410/bj4100213add.htm). They were obtained from Eurogentec (Seraing, Belgium) and some of them have been described previously. The expression of fucosyltransferase (FUT1 to FUT7 and FUT9 to FUT11), sialyl- [ST3GAL3, ST3GAL4, ST3GAL6; ST6GAL1 and ST6GAL2], sulfotransferase- [GAL3ST2 to GAL3ST4 and CHST2, CHST4, CHST6 and I-GlcNAc6ST (GlcNAc-6-O-sulfotransferase gene)] and mucin (MUC4, MUC5AC, MUC5B and MUC6) genes was first studied in control and treated explants by end-point PCR, using an appropriate positive control. For each end-point PCR reaction, amplifications were performed using 2 μl of total cDNA from control or treated explants in a final volume of 50 μl, by using the PCR mix Taq & Go™ Mastermix (Q-biogene; final composition: 10 mM Tris/HCl, pH 9.0, 50 mM KCl, 1.5 mM MgCl2, 0.1% Triton X-100, 0.2 mg/ml BSA, 100 μM each of dNTP, 0.02 units/μl Taq DNA polymerase and 2% glycerol) and 150 μM of sense and antisense primers. When primers were from a previous study, the amplification conditions were according to the author's instructions (see Supplementary Tables 1 and 2 for references). For primers designed for the present study, the amplification conditions were the following: 95 °C for 2 min, 1 cycle; 95 °C for 1 min, Ta (annealing temperature; see Supplementary Table 1) for 1 min, 72 °C for 1 min, 39 cycles, 72 °C for 5 min, 1 cycle. A 20 μl portion of the PCR reaction was submitted to electrophoresis on an agarose gel [2% in TBE (Tris/borate/EDTA) buffer (1×TBE=45 mM Tris/borate and 1 mM EDTA, pH 8.0)] containing ethidium bromide. Gels were photographed under UV light, and the expression of the different enzymes was analysed. This first step allowed us to select the genes expressed in control or treated human bronchial explants.
Variations in the enzyme mRNA levels present in control or treated bronchial explants were then quantified by real-time PCR. Real-time PCR and subsequent data analysis were performed using the Mx4000 Multiplex Quantitative PCR System (Stratagene, La Jolla, CA, U.S.A.) equipped with Version 3.0 software. Each 25 μl PCR reaction contained 12.5 μl of the 2× Brilliant® SYBR® Green QPCR Mastermix (Stratagene), 300 nM of each primer and 2 μl of cDNA diluted 1:20. DNA amplification was performed with the following thermal cycling profile: initial denaturation at 94 °C for 10 min, 40 cycles of amplification [denaturation at 94 °C for 30 s, annealing at Ta (Ta depending on the glycosyltransferase or mucin gene of interest, see Supplementary Tables 1 and 2) for 30 s, and extension at 72 °C for 45 s] and a final extension at 72 °C for 5 min. The fluorescence monitoring occurred at the end of each cycle. The analysis of amplification results was performed using the Mx4000 3.0 software. Threshold cycle value (Ct) is defined as the number of PCR cycles where the fluorescence signal exceeds the detection threshold value. The amplification efficiencies were determined by serial dilution of cDNA synthesis products from a pool of total RNA extracted from bronchial explants or from different cell lines (HT-29, HepG2, HeLa and leucocytes) and were found in the range 97–110%. The specificity of the amplification was checked by recording the dissociation curves after each run, and visualizing the amplified products by 2% agarose-gel electrophoresis in the presence of ethidium bromide. RPLP0 (acidic ribosomal phosphoprotein P0) was used as a control gene to normalize the expression of our genes of interest. To determine the relative amount of the different glycosyl- and sulfotransferases in the treated explants compared with the controls, we first checked the efficiency of the amplifications by making standard curves using serial dilutions of cDNA from bronchial explants or from an appropriate cell line when the expression was too low in the explants. The equation Ct=a×log(initial quantity)+b allowed us to determine the efficiency of the reactions, and to use the Ct obtained in our studies to determine the quantity of transcripts of the different glycosyl- and sulfotransferases.
Variations in mRNA levels of glycosyltransferases were calculated on six individual mucosae treated or not with IL-6 or IL-8 (20 ng/ml, 16 h). The Student's t test was used to compare the means of the relative amount of each glycosyl- or sulfotransferase between the groups of treated and control explants. Statistical significance for this analysis was considered at P<0.05.
Preparation of total proteins from bronchial explants
After a 16 h incubation in CMRL-1066±20 ng/ml of IL-6 or IL-8, pieces of mucosa were collected and centrifuged at 10000 g for 5 min at 4 °C and then rinsed twice in ice-cold PBS. Pieces of mucosa were then disintegrated on ice with a glass homogenizer, in a modified RIPA buffer [150 mM sodium chloride and 50 mM Tris/HCl, pH 7.4, containing 1 mM EDTA, 1 mM PMSF, 1% Triton X-100, 1% sodium deoxycholic acid and a cocktail of protease inhibitors (Roche)]. Protein extracts were then centrifuged at 10000 g for 5 min at 4 °C to get rid of non-homogenized tissue. The concentration of supernatants was determined with the Micro BCA™ Protein Assay Reagent kit (Pierce). The lysates were then stored at −80 °C and subsequently used for immunoblot analysis.
Immunoblot analysis of bronchial proteins for sialyl-Lewisx and 6-sulfo-sialyl-Lewisx determinants
Total proteins (50 μg) from explants treated or not with IL-6 or IL-8 (n=5) were boiled for 5 min in reducing Laemmli sample buffer and resolved by SDS/PAGE on 3–10% gels (15 cm×20 cm). After transfer on to a nitrocellulose membrane (250 mA, 5 h), blocking was performed in TBS (Tris buffered saline) containing 0.05% Tween 20 and 5% (w/v) non-fat dried milk for 90 min. Part of the membrane was treated with C. perfringens sialidase for 16 h at 37 °C as a control for the specificity of the antibodies. Primary antibody incubations with anti-sialyl-Lewisx (clone CSLEX1; BD Pharmingen) or anti-6-sulfo-sialyl-Lewisx (G152, a gift from Dr R. Kannagi; ) antibodies were performed overnight at 4 °C in PBS, 5% dried milk and 0.05% Tween 20, at 1:500 dilution for CSLEX1 or at 1:20 dilution for G152. After three washing steps of 10 min in TBS, 5% dried milk, 0.05% Tween 20, incubation for 1 h at room temperature (20 °C) with HRP-conjugated goat anti-mouse IgG was performed (dilution 1:3000 in TBS, 5% dried milk and 0.05% Tween 20). Membranes were finally washed three times for 5 min in TBS, 0.05% Tween 20 and twice in TBS, and detection was achieved using enhanced chemiluminescence (ECL® Western blotting detection reagents, Amersham Biosciences, Little Chalfont, Bucks., U.K.). MUC4 detection in explants was carried out as previously described using a 1 μg/ml dilution of the antibody m8G7 .
Lectin blot analysis
Sample preparation and electrophoresis were conducted as described above. After transfer on to a nitrocellulose membrane (250 mA, 5 h), blocking was performed in TBS containing 2% polyvinylpyrrolidone for 90 min, followed by incubation with alkaline phosphatase-conjugated MAA or alkaline phosphatase-conjugated SNA, lectins that recognizes preferentially α2,3-linked or α2,6-linked sialic acid residues respectively (dilution 1:75 in TBS containing 1 mM MgCl2, 1 mM MnCl2 and 1 mM CaCl2, pH 7.5) for 90 min. After two washing steps of 10 min in TBS, detection was achieved using NBT/BCIP (Nitro Blue Tetrazolium/5-bromo-4-chloroindol-3-yl phosphate) diluted 1:50 in 0.1 M Tris/HCl (pH 9.5) containing 0.1 M NaCl and 0.05 M MgCl2.
Effect of IL-6 and IL-8 on the expression of sialyl-Lewisx and 6-sulfo-sialyl-Lewisx epitopes on proteins from bronchial explants
In order to determine whether pro-inflammatory IL-6 or IL-8 could participate in the remodelling of the glycosylation pattern of respiratory mucins and increase sulfation and sialylation of Lewisx epitopes in CF patients, total proteins extracted from bronchial explants, treated or not with IL-6 or IL-8 were analysed by Western blotting. Samples from bronchial mucosa from five independent individuals were used to test the effect of both cytokines. α2,3- and α2,6-sialylation were analysed using MAA and SNA lectins respectively, and sialyl-Lewisx and 6-sulfo-sialyl-Lewisx by immunoblotting using specific mAbs. Staining with MAA and SNA, lectins that recognize preferentially α2,3- or α2,6-linked sialic acid residues respectively, was not modified in bronchial samples treated with either IL-6 or IL-8 (Figure 1), suggesting that there was no overall increase in α2-3- or α2-6-sialylation after cytokine treatment. In parallel, the binding of anti-sialyl-Lewisx mAb was increased after IL-6 or IL-8 treatment. Results from two representative patients are shown in Figure 2(A). In all samples, the anti-sialyl-Lewisx mAb labelled proteins of high molecular mass (above 170 kDa). The intensity of the staining was variable between the different control samples, and the number of bands was between 1 and 3, all of high molecular mass and with a diffuse pattern, suggesting that they could correspond to mucins. Strikingly, in all samples containing proteins from explants stimulated by IL-6, the intensity of the signal increased (left panel, explants 1 and 2). The same result was observed for the explants treated by IL-8 (right panel, explants 3 and 4).
Lectin blot analysis of total proteins from human bronchial explants treated or not with IL-6 or IL-8 (20 ng/ml, 16 h)
Immunoblot analysis of sialyl-Lewisx and 6-sulfo-sialyl-Lewisx expression in total proteins from human bronchial explants treated or not with IL-6 or IL-8 (20 ng/ml, 16 h)
Results obtained with the anti-6-sulfo-sialyl-Lewisx antibody are shown in Figure 2(B) (results are shown for two representative patients). In control explants, the anti-6-sulfo-sialyl-Lewisx antibody recognized only one band of high molecular mass (Figure 2B). The intensity of the staining for 6-sulfo-sialyl-Lewisx epitope was strongly increased in explants treated by IL-6 or IL-8, showing that bronchial explants treated with IL-6 or IL-8 contained increased amount of this epitope.
These results suggest that IL-6 or IL-8 did not significantly modify the overall sialylation pattern of bronchial mucosa but only increased the expression of specific glycan motifs, such as sialyl-Lewisx and 6-sulfo-sialyl-Lewisx.
Identification of proteins carrying the 6-sulfo-sialyl-Lewisx epitope
Immunoblotting experiments suggest that proteins that carry sialyl-Lewisx and 6-sulfo-sialyl-Lewisx epitopes in bronchial explants could correspond to mucins, because of their high apparent molecular mass and diffuse migration pattern. In order to test this hypothesis, we quantitatively determined using real-time PCR the level of mucin gene expression in the bronchial mucosa and checked whether their expression was modified by cytokine treatment. In all samples, MUC4 was more expressed than MUC5B, which was more expressed than MUC5AC in most of the patients (Figure 3). MUC6 was expressed at very low levels (Ct>35) in bronchial explants and therefore could not be accurately quantified. The effect of IL-6 or IL-8 on the expression of MUC4, MUC5AC, and MUC5B in bronchial explants was then investigated and quantified using real-time PCR. Our results clearly showed that there was no consistent effect of these cytokines on the expression of MUC4, MUC5AC and MUC5B mRNA in human bronchial explants. Important inter-individual variations in the level of transcripts were observed under control conditions and the effect of IL-6 and IL-8 was extremely variable between samples, with sometimes a stimulatory effect, or an inhibitory effect, or no effect at all (results not shown). These results indicate that the increase in sialyl-Lewisx and 6-sulfo-sialyl-Lewisx epitopes cannot be directly linked to increased expression of mucin genes. The presence of mucins in the bronchial samples was also analysed in total protein extracts from bronchial explants. Immunoblotting experiments showed a high expression of MUC4 in bronchial protein extracts. MUC1 and MUC5B were also detected but at a lower level compared with MUC4, and MUC5AC was not detected, confirming the results of real-time PCR analysis (results not shown, and Figure 3). Moreover, the band of high molecular mass visualized using the 6-sulfo-sialyl-Lewisx antibody was in most cases also stained with the anti-MUC4 antibody (Figure 4), but not with anti-MUC5B or anti-MUC1 mAbs (results not shown), suggesting that MUC4 is the main protein carrying 6-sulfo-sialyl-Lewisx epitopes in the human bronchial mucosa.
MUC4, 5AC and 5B mRNA in control human bronchial explants ( n=8)
Immunoblot analysis of 6-sulfo-sialyl-Lewisx and MUC4 expression in total proteins from human bronchial explants treated or not with IL-6 or IL-8 (20 ng/ml, 16 h)
Effect of IL-6 and IL-8 on the transcriptional expression of sialyl-, fucosyl- and sulfotransferases from bronchial explants
In order to determine which enzymes are responsible for the increased amounts of sialyl-Lewisx and 6-sulfo-sialyl-Lewisx found on human bronchial proteins in IL-6- or IL-8-treated explants, we first made a survey of fucosyl-, sialyl- and sulfotransferases expressed in human bronchial explants. We analysed the expression of enzymes potentially involved in the biosynthesis of sialyl-Lewisx and 6-sulfo-sialyl-Lewisx epitopes, i.e. α1,3/4-fucosyltransferases (FUT3 to FUT7 and FUT9 to FUT11), α2,3-sialyltransferases (ST3GAL1 to ST3GAL4 and ST3GAL6) and GlcNAc-6-O-sulfotransferases (CHST2, CHST4, CHST6 and I-GlcNAc6ST). In addition, because human bronchial mucins were shown to contain Galβ1-4GlcNAc disaccharide units substituted with α1,2-linked fucose residues, α2,6-linked sialic acid residues, or sulfate on the C-3 of the Gal, we also investigated the expression of α1,2-fucosyltransferases (FUT1 and FUT2), α2,6-sialyltransferases (ST6GAL1 and ST6GAL2) and Gal-3-O-sulfotransferases (GAL3ST2, GAL3ST3 and GAL3ST4). After screening by end-point PCR, the expression of enzymes that were detected in control or treated explants was then analysed by real-time PCR. Our results indicate that ST6GAL1 was more expressed than ST6GAL2 in all control bronchial samples (in average 40-fold more) and that the expression of ST6GAL1 was not affected by the addition of cytokines, whereas the expression of ST6GAL2 was highly stimulated by IL-6 or IL-8, with 5.7–46-fold or 5–11.8-fold increases respectively (Figures 5A and 5B). Among the ST3GAL genes studied, ST3GAL4 was the most expressed gene in bronchial explants (the ratios of absolute expression levels of ST3GAL4/ST3GAL3 and ST3GAL4/ST3GAL6 were 6.3 and 10.8 respectively). No significant variations in the transcriptional expression of α2,3-sialyltransferases ST3GAL3 and ST3GAL4 was observed when IL-6 or IL-8 was added, whereas the addition of IL-6 significantly increased the expression of ST3GAL6 in all the bronchial explants tested (Figure 5A). Among the FUT genes studied, FUT1, FUT4, FUT5, FUT6, FUT7 and FUT9 were not expressed in the bronchial explants (control or treated with cytokines). On average, FUT10 was the most expressed FUT in the control bronchial explants (FUT10/FUT3 and FUT10/FUT11 ratios were 2.5 and 4.2 respectively). Only FUT3 (encoding the Lewis α1,3/4-fucosyltransferase) showed an increased expression after IL-8 treatment, whereas the expression of FUT11 (encoding a potential fucosyltransferase of still unknown specificity) was moderately but significantly increased in the different explants tested (Figures 5A and 5B). GAL3ST2 and GAL3ST3 genes encoding Gal-3-O-sulfotransferases were not expressed at quantifiable levels in the bronchial explants, and the expression of GAL3ST4 was not significantly affected by the addition of IL-6 or IL-8 (Figure 5). Among the GlcNAc-6-O-sulfotransferase-encoding genes, only CHST4 and CHST6 were expressed in our model, with CHST6 in average expressed 2.8-fold more than CHST4. Both enzymes had their expression increased under IL-6 treatment, whereas IL-8 had only a significant stimulatory effect on CHST6 expression (Figure 5). Altogether, these results identify the glycosyl- and sulfotransferases that could be involved in the increased expression of sialyl-Lewisx and 6-sulfo-sialyl-Lewisx epitopes in the bronchial mucins in inflammatory mucosae of CF patients (1).
Effect of IL-6 (A) and IL-8 (B) (20 ng/ml, 16 h) treatments on the relative expression of fucosyl- (
FUT2, FUT3, FUT10 and FUT11), sialyl- ( ST3GAL3, ST3GAL4, ST3GAL6, ST6GAL1 and ST6GAL2) and sulfotransferases ( GAL3ST4, CHST4 and CHST6) in human bronchial explants
Glycosyl- and sulfotransferases that are up-regulated by the addition of IL-6 or IL-8, potentially involved in the increased expression of sialyl-Lewisx and 6-sulfo-sialyl-Lewisx epitopes in the bronchial mucins of CF patients
Our study clearly demonstrates the influence of IL-6 and IL-8, two pro-inflammatory cytokines, on the expression and activity of several fucosyl-, sialyl- and sulfotransferases involved in the biosynthesis of sialyl-Lewisx and 6-sulfo-sialyl-Lewisx epitopes by the human bronchial mucosa. Several independent studies have previously shown that IL-6 and, to a lesser extent, IL-8 can modulate protein secretion and glycosylation [23,24]. IL-6 also modulates the expression of some mucin genes by inducing an early response of MUC2, MUC5B and MUC6 mucin genes in LS180 cells . In parallel, IL-8 is a chemokine that attracts neutrophils to the inflammatory sites and seems to be constitutively up-regulated in CF cells . Its effect on glycosylation is not well documented but IL-8 has been shown to stimulate the expression of MUC5AC and MUC5B genes . Considering that these two cytokines are key actors in the chronic and destructive inflammatory response exhibited by the lung of CF patients, it is therefore of particular interest to investigate the effect of these two pro-inflammatory cytokines on glycosylation processes in human bronchial cells.
Our results show that the increase in sialyl-Lewisx and 6-sulfo-sialyl-Lewisx in cytokine-treated explants is specific to a very limited number of high-molecular-mass proteins, whereas staining with MAA or SNA reveals a broad population of α2,3- and α2,6-sialylated glycoproteins unaffected by cytokine treatment. These results are in agreement with those of Davril et al.  who have shown that the amount of sialyl-Lewisx epitopes on high-molecular-mass bronchial glycoproteins depends on the severity of airway infection/inflammation and on the nature of the bacteria present in the sputum. The influence of inflammation on the biosynthesis of 6-sulfo-sialyl-Lewisx epitope has not been studied before, and it is the first time that a stimulatory effect of pro-inflammatory cytokines on that biosynthetic pathway has been shown. The anti-6-sulfo-sialyl-Lewisx labelling showed a broad band of high molecular mass, and the immunostaining with anti-MUC mAbs strongly suggests that this epitope is carried by the membrane-bound MUC4.
Variations in the levels of expression of MUC genes were observed between the bronchial samples, which is not surprising considering the different factors inducing heterogeneity in the model of human bronchial explants. Indeed, explants were obtained from patients differing in age, sex, stage of the disease and treatment before surgery, and the lung area from which the explant was obtained varied. It is known that the gel-forming mucin genes (MUC2, MUC5AC, MUC5B and MUC6) are regulated at the transcriptional level by pro-inflammatory cytokines such as IL-1β, TNFα and IL-6, pleiotropic cytokines (IL-4, IL-13 and IL-9), as well as by other mediators like growth factors and hormones [25,28,29]. In the present study, the effect of IL-6 and IL-8 on MUC4, MUC5AC and MUC5B expression was investigated, but inconsistent effects of both cytokines on the expression of these genes was observed, with sometimes opposite effects from one sample to the other. This seems to indicate that the effect of cytokines on MUC gene expression observed with cultured cells is not directly transposable to bronchial explants.
As it has been previously reported by Delmotte et al. , no expression of FUT genes FUT1, FUT5, FUT6 and FUT7 was observed in bronchial explants. In addition, we found that FUT9 gene encoding an enzyme preferentially synthesizing the Lewisx epitope  is not expressed in the control or treated bronchial explants and FUT4, encoding a fucosyltransferase that preferentially synthesizes the sialyl-Lewisx epitope , was only detectable in a few samples. Interestingly, FUT10 and FUT11 genes, encoding putative α1,3/4-fucosyltransferases of still unknown activities , are expressed in human bronchial mucosa. As already described, human bronchial explants express FUT2, encoding the secretor enzyme, and FUT3, encoding the Lewis enzyme . The result of our screening on fucosyltransferase expression is in agreement with the analysis of FUT genes in mouse lung by Comelli et al.  using a glycogene focused microarray approach. Quantitative PCR analysis of FUT2, FUT3, FUT10 and FUT11 transcripts in treated explants revealed that only FUT3 and FUT11 expression was significantly modified, after treatment with IL-8 and IL-6 respectively. The α1,3/4-fucosyltransferase FucT III is, together with FucT IV and FucT VII, involved in the sialyl-Lewisx biosynthesis pathway and the increased expression of FUT3 could be therefore responsible for the increased synthesis of the sialyl-Lewisx epitope on human bronchial glycoproteins. This is in agreement with a recent report showing the increased expression of FUT3 and sialyl-Lewisx in TNF-α in the mucoepidermoid lung cancer cell line NCI-H292 . It is also of particular interest that FUT11 is expressed and up-regulated by IL-6 in the human bronchial explants. The FUT11 gene encodes a Golgi-localized protein of still unknown enzymatic specificity. It has been shown that this protein does not fucosylate Galβ1-4GlcNAc units, but its activity has not been tested on other substrates yet. It is therefore possible that FucT XI is a fucosyltransferase involved in sialyl-Lewisx and/or 6-sulfo-sialyl-Lewisx biosynthesis.
The two α2,6-sialyltransferase genes ST6GAL1 and ST6GAL2 are expressed in human bronchial explants and respond in a different way to the addition of cytokines. The expression of ST6GAL1 was not affected by the addition of IL-6 or IL-8, whereas ST6GAL2 was strongly induced by both cytokines in all treated explants. ST6Gal I expression was previously shown to be IL-6-dependent in hepatocytes. In liver, ST6GAL1 is expressed as a 4.3 kb transcript under the control of a liver-specific promoter different from the promoter used in other tissues, in which ST6GAL1 is expressed as a 4.7 kb transcript . This tissue-specific expression of ST6GAL1 could explain the absence of induction in cytokine-treated explants. Whereas ST6GAL2 was strongly induced in all cytokine-treated explants, SNA labelling shows that α2,6-sialylation of glycoproteins is not modified in treated bronchial explants. This result could be explained by the differences in substrate specificity and expression levels of these two enzymes: ST6Gal I is an ubiquitously expressed enzyme involved in the α2,6-sialylation of glycans carrying type 2 disaccharide units (Galβ1-4GlcNAc). ST6Gal II shows a much more restricted expression pattern and exhibits a slightly different substrate specificity [36,37], acting preferentially on LacdiNAc structures (N,N′-diacetyl-lactosediamine; GalNAcβ1-4GlcNAc), a motif present on very few human glycoproteins. It is likely that the α2,6-linked sialic acids present in explants (and visualized by SNA staining) are carried by the classical type 2 disaccharides present mostly on N-glycan branches, which are a better substrate for ST6Gal I. In addition, although not induced by cytokines, the level of expression of ST6GAL1 is higher than ST6GAL2 (even in treated explants). Until now, sialyl-LacdiNAc has never been characterized in human bronchial tissue but its limited expression on to specific glycoproteins could occur and play a specific role during inflammation.
Among ST3GAL genes (ST3GAL3, ST3GAL4 and ST3GAL6) involved in the sialylation of type 2 units (the precursor of the sialyl-Lewisx epitope), only ST3GAL6 expression was significantly increased in IL-6-treated bronchial explants. This enzyme was previously shown to be expressed in human and mouse lungs [33,38] and is therefore a good candidate for the synthesis of the sialyl-Lewisx epitope in this tissue. ST3GAL3 and ST3GAL4 are both expressed in our model but their expression was not significantly affected by IL-6 or IL-8.
Human bronchial mucins were shown to contain sulfate residues, either on the C-3 of terminal Gal residues, or on the C-6 of internal GlcNAc residues . In CF, bronchial mucins are highly sulfated, and the structures determined by 1H-NMR and MS suggest that CF bronchial mucins carry more sulfated GlcNAc residues than mucins from patients suffering from chronic bronchitis. Genes encoding Gal-3-O-sulfotransferases active on glycoproteins showed either very low or no expression in our model, except GAL3ST4. The sulfotransferase Gal-3-O-sulfotransferase was previously shown to be active on mucin-type structures . Although Delmotte et al.  could detect an increased Gal-3-O-sulfotransferase activity in TNFα-treated bronchial explants, the addition of IL-6 or IL-8 did not significantly modify the expression of ST3GAL4 in a similar model.
Among genes encoding GlcNAc-6-O-sulfotransferases, we studied CHST2 and CHST4, two genes encoding enzymes able to synthesize the 6-sulfo-sialyl-Lewisx epitope, especially on HEV (high endothelial venules). CHST2 was not expressed in bronchial explants, whereas the expression of CHST4 was slightly increased in the presence of IL-6. The expression of CHST4 was previously shown to be induced by TNF-α in cultured endothelial cells . It is of interest that both allergic sheep and human asthmatics exhibit airway secretions, presumably mucins that are stained by the mAb MECA-79 . MECA-79 recognizes mucin-type core-1 extended with a 6-sulfo-sialyl-Lewisx determinant. It seems that the major source of MECA-79-reactive glycoproteins is the airway epithelium in which CHST4 was found to be expressed. It has been proposed that the MECA-79 reactive secretions could provide extravascular ligands (adhesive or signalling) for leucocytes migrating into airways .
We also investigated variations in the expression of genes encoding sulfotransferases acting on glycoproteins but not preferentially involved in 6-sulfo-sialyl-Lewisx biosynthesis, i.e. the gene encoding GlcNAc6ST3 , and CHST6  that acts preferentially on terminal GlcNAc residues of keratan sulfate chains. Enzymes encoded by these two genes have not been yet implicated in 6-sulfo-sialyl-Lewisx biosynthesis, but they both sulfate terminal GlcNAc residues of oligosaccharide chains and are therefore potential candidates for this biosynthesis pathway. The gene encoding GlcNAc6ST3 was not expressed in either control or treated bronchial explants. CHST6 expression was induced both by IL-6 and IL-8 in the bronchial explants, although only moderately by IL-8. Our results point at CHST4 and CHST6 as major sulfotransferase genes expressed and induced by cytokines in bronchial explants. Altogether, the present study allows us to suggest which enzymes are involved in the cytokine-induced sialyl-Lewisx and 6-sulfo-sialyl-Lewisx biosynthesis in the bronchial mucosa of CF patients (1).
Sialyl-Lewisx epitopes have been detected on mucin-type glycoproteins in different cell types and display important biological roles. Cell surface sialyl-Lewisx is recognized by the endothelial selectins, is involved in leucocyte homing and rolling during inflammation . Sialyl-Lewisx is also overexpressed in many carcinomas and is usually considered as a poor prognosis factor, since the interaction of this epitope with selectins facilitates extravasation and spreading of cancer cells . Interactions between macromolecules carrying abnormally expressed Sialyl-Lewisx and cells carrying selectins (leucocytes and endothelial cells) could modulate the inflammatory response, by competing with the relevant selectin ligands . 6-Sulfo-sialyl-Lewisx is a major sulfated L-selectin ligand present on mucin-type proteins such as CD34 and GlyCAM-1 (glycosylation-dependent cell adhesion molecule 1) on HEV of lymph nodes and is essential for lymphocyte rolling and lymphocyte recruitment . Our results suggest that in airway mucosa, the inflammation-induced 6-sulfo-sialyl-Lewisx determinant is carried by the membrane-bound bronchial mucin MUC4. Interestingly, MUC4 overexpression is a poor prognosis factor in pancreas and lung cancers, and MUC4 has been shown to be involved in pancreatic cancer progression, by modifying interactions between cancer cells and the extracellular matrix . Importantly, MUC4 expression is negatively regulated by CFTR, the protein deficient in CF patients . In the bronchial mucosa of severely infected CF patients, one could therefore expect increased levels of abnormally glycosylated MUC4 that could affect signalling and interactions. The increased expression of sialyl-Lewisx and 6-sulfo-sialyl-Lewisx after cytokine treatment in human bronchial explants is therefore of important biological relevance and offers an increased number of potential ligands for leucocytes and for the binding of P. aeruginosa, and may therefore contribute to the chronicity of airway infection in CF. Moreover, the interactions of P. aeruginosa with mucins may prevent its opsonophagocytic killing by human PMN cells (polymorphonuclear cells) .
Here, we show which glycosyl- and sulfotransferases are up-regulated by IL-6 and IL-8 in the human bronchial mucosa, resulting in increased amounts of sialyl-Lewisx and 6-sulfo-sialyl-Lewisx epitopes on bronchial proteins, which are key determinants in the development and chronicity of the lung inflammation and infection. In the future, the study of the mechanisms by which these enzymes are regulated in the bronchial mucosa could offer new possibilities to interfere with the development of the disease in CF patients by focused approaches targeting bronchial glycosyl- and sulfotransferases by specific inhibitors or compounds that modulate their expression or activity.
We are grateful to Dr R. Kannagi for providing us with the anti-6-sulfo-sialyl-Lewisx antibody G152. We are also indebted to Dr Eric Mensier (Polyclinique La Louvière, Lille, France), Dr J. M. Faillon (Clinique du Bois, Lille, France) and Dr M. Debaert (Polyclinique La Louvière, Lille, France) for providing us with human bronchial explants. This work was supported by ‘Vaincre la mucoviscidose’.
CF transmembrane conductance regulator
carbohydrate sulfotransferase gene
high endothelial venules
Maackia amurensis agglutinin
acidic ribosomal phosphoprotein P0
Sambucus nigra agglutinin
Tris buffered saline
tumour necrosis factor